Note: In these early versions, this document will be changing rapidly. If the document you are holding is dated more than 1 month ago, please obtain a new copy of these notes, for inaccuracies and omissions surely abound. Thanks!

	September 27, 1999 Pittsburgh
Dear RI Tour Guide, Until very recently, the only resource for giving tours was an excellent set of notes provided by Henry Schneiderman. His notes were for specific stops out the tour, and have been incorporated into this guide. So, before I begin, I would like to thank Henry for allowing me to copy his notes verbatim. Of course, any errors in this document, even in parts where I mention Henry, are my own fault an no one else's. VERY BIG DISCLAIMER: I put these notes together before the sample tour that I gave in the Fall of 1999, and then revised them before the same in the fall of 1999. The information in it is mostly from memory, though some of the statistics about the Institute were culled from the newly remodeled RI web pages. Therefore, while I have made every effort to be factually correct, if you notice any errors, please inform me immediately. EVEN BIGGER WARNING: Because of the scope of the information in this document, it is really easy for information to become out-of-date. If you notice anything that is no longer correct, please let me know so that I can update it or remove it. LITTLE REQUEST: If you have any information to add to what I have here, whether it be neat anecdotes that help a particular tour stop, statistics that demonstrate a point, other non-tour-stop information, or even extra tour stops, please let me know so that I can add it to this page. If you're looking for inspiration, see Missing Topics, below. If you have any comments or suggestions, feel free to send me e-mail, call me, or even, gasp, stop me in the hall and chat. See you around.
	Yours, Sal Desiano

• Descriptions of some seminars
• Descriptions and locations of labs, centers, and groups in RI Structure
• Descriptions of RI buildings and facilities (see Of Interest)
• Information on SCS construction
• Better story of the RI founding
• Information on Porter Hall laboratories
• Information on Doherty Hall laboratories
• Information on Hamburg Hall laboratories
• Anecdotes, stories, information from institutional memory

These are tidbits and items that might come up during your tour, about which you can talk while walking from one place to another, or about which you might be asked. They aren't associated with any particular location, though a bit of About the RI or History of Robotics might be appropriate for the beginning of a tour.

About the RI
The RI was founded in 1979 by a group of people (about 9) in SCS. I don't remember who these people are. The idea was to have a centralized location in which fundamental issues of robotics (virtually none of which were solved back then) could be studied. From the RI Web Page,

The Robotics Institute at Carnegie Mellon University was established in 1979 to conduct basic and applied research in robotics technologies relevant to industrial and societal tasks. Seeking to combine the practical and the theoretical, the Robotics Institute has diversified its efforts and approaches to robotics science while retaining its original goal of realizing the potential of the robotics field.

The RI is one of the largest research institutions in the world dedicated to the study of robotics. There are only a few that are possibly larger than us, though I don't know what they are. People have made noises about somewhere in St. Petersburg and somewhere in China.

We also have one of the most wide-ranging and varied robotics research agenda, ranging from wheeled robotics to manipulation to speech to manufacturing.

Currently (as of 9/98), the RI has more than 331 people, in addition to graduate students from other departments that work in our labs. The 331 we can count can be broken down into the following categories:

53	Faculty
4	Adjunct Faculty (big names, not necessarily here)
14	Post Doctoral Fellows
33	Visitors (visiting for various terms as professors, scientists, researchers, or students, from various institutions and corporations around the world)
19	Support Staff
67	Ph.D. Students
2	Full-time Masters Students
16	Part-time Masters Students
40	Undergraduate students (employed in various labs, or getting involved in our undergraduate program)

In 1998, we had a total research budget of $30 million (this information is on our web page, though there's no particular reason to announce this on a tour). Again, unless you're pitching the RI to a particular group, there's no particular reason to announce where our funding comes from, but it is public knowledge that our funding comes from:

• NASA (National Air and Space Administration)
• NSF (National Science Foundation)
• DARPA (Defense Advanced Research Projects)
• DOT (Department of Transportation)
• DOE (Department of Energy
• US and Foreign Industry

A Very Brief History of Robotics
As with much of the work we do in Robotics today, the science we produce is predicted and predated by the science fiction that we read. The word "robot" comes from a play called "Rossum's Universal Robots", written by Karl Capek in 1926. Throughout the 1920s and 1930s, robots appeared in science fiction stories, primarily in negative contexts. In 1941, Isaac Asimov coined the term "robotics" as the study of robots, and produced the first in a wave of stories about robots that helped and aided their creators.

The technical development of robotics happened simultaneously in two areas of research. In the 1950s, the manufacturing industry began developing more and more flexible machines to use in their factories. In the late 1950s, the Unimate Company produced the first robot arm, and it was marketed toward manufacturing. From this work came the study of manipulation, manipulators, industrial robotics, and related fields.

In the 1940s, people in academia were working on teleoperation, artificial intelligence, image processing, and natural language. Over time, these fields merged together under the auspices of Artificial Intelligence. In 1969, AI researchers (primarily people working on planning) got the itch to try their work on real systems, and the field of mobile robotics was born.

In the 1970s, mobile robotics, manipulators, industrial robotics, and the developing field of machine vision began to flow together, and the field of robotics began to take shape. It was at this point that Carnegie Mellon began work in robotics, and the diverse origins of the field that made the founding of the Robotics Institute so opportune.

RI Structure
This comes up in a question once in a while, but you don't need to bring it up if nobody asks (particularly because it's absurdly complicated). The hierarchical structure at CMU is kind of confusing. In general, the university is broken down into seven "colleges" or "schools": The School of Computer Science, the Carnegie Institute of Technology, the College of Fine Arts, the College of Humanities and Social Sciences, the Graduate School of Industrial Administration, the H. John Heinz III School of Public Policy and Management, and the Mellon College of Science. There is no distinction between a "college" and a "school".

Under each school, there are "departments" and "institutes". In general, "institutes" have more autonomy than "departments", and can sometimes be under more than one school. The Robotics Institute is in the School of Computer Science, as is the Computer Science Department (the undergraduate CS program), the Human-Computer Interaction Institute, and the Language Technologies Institute. The Institute for Complex Engineered Systems is under the Department of Electrical and Computer Engineering, which is under the Carnegie Institute of Technology. The Software Engineering Institute isn't under any college or school at all.

Under all of this are the "centers", which vary in autonomy, size, and hierarchy. For the most part, centers are under a particular department or institute. For example, the Field Robotics Center is under the Robotics Institute. Wierd cases abound, however, like the former Engineering Design Research Center that was under both the RI and ECE.

Finally, "laboratories" are generally within a particular department or institute, and can be within a particular center. The "Manipulation Laboratory" is directly under the Robotics Institute, while the "Mobile Robot Programming Laboratory" is under the "Vision and Autonomous Systems Center", which is under the "Robotics Institute".

Got it? Okay, here are most of the groups related to the RI, organized as best as I can do without a three-dimensional graph. I'm sure I've missed a few, so just let me know and I'll add things. I tried to include anything that might come up on a tour (so, notably, there are very few details regarding the REC or ICES on this list),

• School of Computer Science (a.k.a. SCS)
• Robotics Institute
• Center for Integrated Manufacturing and Decision Systems (a.k.a. CIMDS)
• Case Based Reasoning Laboratory
• Distributed Multi-Agent Systems
• Electronic Commerce
• Evolutionary Computation
• Intelligent Coordination and Logistics Laboratory
• Intelligent Sensor, Measurement, and Control Lab
• Rapid Manufacturing Laboratory
• Visualization and Intelligent Interfaces Group
• Field Robotics Center (a.k.a. FRC)
• Medical Robotics and Computer Assisted Surgery (a.k.a. MRCAS)
• National Robotics Engineering Consortium (a.k.a. NREC)
• Vision and Autonomous Systems Center (a.k.a. VASC)
• Advanced Mechatronics Lab
• Face Group
• Automated Highway Systems Group (a.k.a. AHS)
• Mobile Robot Programming Lab
• Video Surveillance and Monitoring (a.k.a. VSAM)
• Vision for Virtual Environments (virtual reality) group
• VLSI Computation sensor laboratory
• Laboratories not under a Center
• Calibrated Imaging Laboratory
• Computer Graphics Laboratory
• Helicopter Laboratory
• Imaging Systems Laboratory
• Interactive Systems Laboratory
• Manipulation Laboratory
• Microdynamic Systems Laboratory
• Microelectromechaincal Systems Laboratory (a.k.a. MEMS)
• Mobile Robot Laboratory
• Parallel Computer Vision Laboratory
• Project Listen
• Robotics Education Laboratory
• Robot Learning Laboratory
• Robotic Sensor Based Planning Laboratory
• Shape Deposition Manufacturing Laboratory
• Tissue Engineering Laboratory
• Carnegie Mellon Research Institute (a.k.a. CMRI)
• Entertainment Technology Center
• Human-Computer Interaction Institute (a.k.a. HCII)
• Language Technologies Institute (a.k.a. LTI)
• Center for Machine Translation (a.k.a. CMT)
• Carnegie Institute of Technology (a.k.a. CIT)
• Department of Electrical and Computer Engineering
• Institute for Complex Engineered Systems (a.k.a. ICES)
• Department of Mechanical Engineering
• Engineering Design Research Center (a.k.a. EDRC)
• Software Engineering Institute
• Other
• Department of Art
• Information Technology Center (a.k.a. ITC)
• Center for the Neural Basis of Cognition (a.k.a. CNBC)

RI Corporate Relations
From the RI web site,

Companies can participate in the activities of the Robotics Institute either by becoming an affiliate of one or more of the research programs, or by sponsoring specific research projects of particular interest to them. Companies that participate with the Institute are given research results before public release allowing sponsors and affiliates great benefit from the research. This rapid technology transfer enables industry to maintain a technological advantage in manufacturing and productivity. Other benefits to participating companies include access to research personnel, information on graduating students, and access to the Institute's extensive library of videotapes describing various research activities.

Similarly, VASC offers a "Corporate Affiliation" program, for which Stephanie Riso is the contact.

RI Facilities and Buildings
We are large department, primarily concentrated in Smith, FRC, PRB, Wean, and Building D. There are labs, however, in many buildings around campus, most of which are listed here. For computer support, we go through, and have access to the resources of, the School of Computer Science.

Doherty

a.k.a. Doherty Hall, a.k.a. DH

Hamburg

a.k.a. Hamburg Hall, a.k.a. HBH, not HH

Porter

a.k.a. Porter Hall, a.k.a. PH

Newell-Simon Building

(Newell-Simon doesn't have a building abbreviation, since it's not on any maps yet. all of these aliases are actually aliases of the two buildings that make up Newell-Simon) a.k.a. Field and Mobile Robotics, a.k.a. Bureau of Mines Building C, a.k.a. FRC, a.k.a. BOM-C, a.k.a. FMR a.k.a. BOM-D, a.k.a. Mine Safety Laboratory, a.k.a. Bureau of Mines Building D, a.k.a. MSL

Planetary Robotics

a.k.a. Planetary Robotics Building, a.k.a. Bureau of Mines Building B, a.k.a. PRB, a.k.a. BOM-B, a.k.a. PRB

Smith

a.k.a. Smith Hall, a.k.a. Elliott Dunlap Smith Hall, a.k.a. EDSH

Wean

a.k.a. WeH, a.k.a. Wean Hall

Historical Note: Until very recently, the Newell-Simon building was actually two separate buildings, BOM-D and Field Robotics. Through 1998-2000, they were totally renovated and combined (see "The Current Construction", below). Before that, they were actually the government offices that studied mines and mine safety, particularly because there are so many mines, now mostly inactive, in the Pittsburgh region. There is still an old set of railroad tracks that run into the car barn in the sub-sub-basement of Building D (the side of Newell-Simon that is closer to Wean):

Field Robotics (now part of Newell-Simon)

a.k.a. Field and Mobile Robotics, a.k.a. Bureau of Mines Building C, a.k.a. FRC, a.k.a. BOM-C, a.k.a. FMR

Bureau of Mines (now part of Newell-Simon)

a.k.a. BOM-D, a.k.a. Mine Safety Laboratory, a.k.a. Bureau of Mines Building D, a.k.a. MSL

We also have various off-campus testing sites for field robotics, and one center is located entirely off campus:

REC

a.k.a. Robotics Engineering Consortium, a.k.a. NREC, a.k.a. National Robotics Engineering Consortium

Located in Lawrenceville

The Current Construction
Over the past year, and through the next two years, the School of Computer Science has been and will be expanding. By now, BOM-D and FRC have had two floors added to each of them, the space between them is filled, there is a third-floor entrance into the Smith parking lot, and there is a bridge to the fourth floor of Wean Hall.

In order to keep the loss of office space to a minimum (during construction), the two new floors were being built first, and then the bottom two floors will be gutted and renovated. Pretty cool? Yes. Violates laws of physics? I think so.

Seminars
One of the strengths of the Robotics Institute is the wealth of original, state-of-the-art research being done across the University and surrounding colleges. Through the strength of our local faculty and students, RI people have access to a wide variety of seminars presented by the people advancing and shaping virtually every field of engineering related to robotics. Following is a list of related seminars that I drummed up on a few minutes notice (note that I don't have descriptions for a few of these, so feel free to pitch in). There are other seminars that I don't know about, so if you think there is a seminar that should be on this list and isn't, please let me know. The dates for each of these seminars are subject to change, and the best way to learn about these seminars are through their web pages.

Robotics Seminar

Our weekly seminar, every Friday, brings in speakers from the CS department, the RI, the ECE department, the ME department, and from similar departments across the country and world.

AI Seminar

Held almost every Tuesday and organized by the CS Department, this seminar series focuses on ground-breaking research in the field of AI, both from CMU and other world-leading universities.

CNBC Talks

The CNBC talks bring in faculty and researchers from around the country and within CMU to talk about neural biology and neural science. These talks are fairly often (every few days), but extremely irregular.

CORAL Seminars

Held on Wednesday afternoons, these talks cover issues relating to software and hardware intelligent agents.

Computer Systems Seminars

This seminar series is hosted by the School of Computer Science and the Electrical and Computer Engineering Department at Carnegie Mellon University. The seminars cover a broad range of topics related to computer systems research.

ECE Seminar Series

As general as the Robotics Seminar, the ECE Seminar Series cover any and all issues relating to electrical and computer engineering. These seminars are not always relevant to robotics, but they often are. These talks are held every Thursday afternoon, and videos of past seminars are often available on the web.

Field Robotics Center Seminar Series

The FRC Seminar Series exists to serve the interests of researchers in FRC. Topics from mobile robots, field robotics applications, sensing, locomotion, control, planning, etc. investigated by researchers from all over are be presented to the community. These seminars run every Wednesday afternoon.

HCII Seminar

To keep abreast of the latest research and the work of our colleagues, the HCII holds a formal seminar series, plus a number of informal meetings.&

SCS Distinguished Lecture Series

Monthly lectures by distinguished members of the computer science corporate and academic communities.

Machine Learning and Friends

Sponsored by CALD, the ML lunches discuss both the practical and theoretical aspects of machine learning. Topics covered include machine learning in general, statistical artificial intelligence, statistical learning theory, text learning, robot learning, and any other machine learning related topics.

VASC Seminars

Offered, usually, on Monday afternoons, these seminars cover issues either relavent to or interesting for researchers in vision and autonomous systems. They tend to be about methods for and implementations of computer vision.

Center for Automated Learning and Discovery Seminar

Needs to be written.

Science of Learning Research Seminar

Needs to be written.

System Design and Implementation Seminar Series

Systems Design and Implementation (SDI) seminar series is held in 8220 Wean Hall on Thursdays between 12 noon and 1:30 pm. Attendance is open to all interested parties. The SDI seminar series is an informal gathering of the SCS systems community over lunch and a talk. The range of talks is broad -- we are interested in the structure, implementation, and performance of substantial systems. Typically we have talks on operating systems, application systems, communications and language implementations. To preserve the informal community nature of this seminar series, most talks are given by CMU researchers, and advertised with only a days warning through a mailing list. Accordingly, most talks do not appear on the schedule until the day of or after their occurrence. Interleaved in this seminar series are a smaller number of talks given by special guests. These talks are given broader advertisement and longer time periods.

---

Giving Tours

Giving tours is a lot like teaching a great class: the students are really interested, and you only need to present the most interesting material. Unlike teaching, the age of your students varies widely, you may not be very familiar with the material you are presenting, and there's a lot of walking involved. With these things in mind, here are a few bits of information that might make your life easier when giving tours.

When to Give Tours
The tour policy dictates who has to give a tour and when. It is published under the RoboOrg web pages at http://www.ri.cmu.edu/roboorg/semi-private/tours. The "and when" portion of that sentence is governed by the tour queue, governed by the tour policy and the standings of which can be found at http://www.cs.cmu.edu/afs/cs.cmu.edu/user/curmson/pub/tq/tourQueue. Since you can be "bumped" into giving a tour, if you see a tour that sounds like a group of people you'd like to show around (you have similar interests, they are from or company or university where you might want to work, they are guests of your advisor, they are an age group that you like interacting with, etc.), then take that tour before you're bumped into giving a tour to a group of ninety year old paleontonlogists from Wales that only speak Gaelic.

The Day Before the Tour
Like teaching, however, there are three things things that always makes your life easier: preparation, preparation, and, oh, yes, preparation. There are a few things that you must do the day before tour, and a few things that you can do before the tour to toughen yourself against Murphy and his famous law.

Things you must do:

• Pick up the tour keys from Suzanne. The correct keys are the one with the GSO keychain attached; the other rings are missing a key, though I'm not sure which one.
• Select your tour stops. This really belongs in the "reccommended" category, but you need to choose them at least a day ahead of time so that you can
• Alert the labs that you are visiting. Each laboratory listed as a tour stop has a contact person that you should alert to your tour. Some labs will have other things going on and won't be able to accommodate your tour when you intended to give it. Yes, accommodate. While it is to a labs advantage that they receive publicity, they may have too many things to do. Would you give a tour of your mother's house without warning her?

Things you should do:

• Visit the labs, or who knows what will happen when you go there during the tour. You could combine this with alerting the lab to your tour. Make sure that the keys work any that any props you were intending to use are there. Cue up any videos that you are going to use -- it's really hard to fill the dead time while you try to cue up the video during the tour.
• Plan your route. For obvious reasons. Take into account the kind of people that you are taking around (see below). Remember that you're probably going to have to bring the tour back to where you started it, so work that into the route unless you want a long, nervous silence on the walk back from Wean to Smith.
• Scan the tour notes though you've, of course, read them from end to end already, you're less likely to be caught flat-footed if the material is fresh in your head.

During the Tour
No matter what kind of tour you are giving, there are a number of things that they have in common. It often helps if you start the tour with an overview of the Institute, robotics as a whole, or the center on whose research you are focusing. This gives people a context in which to file all the things that they are about to see. There are two excellent places to start tours: the lobby of Smith, and the lobby of FRC. Both lobbies have nice, large displays of about a dozen of our robots and projects. These can serve as an excellent opportunity to discuss projects that you won't see on your tour, either because they are out of the way, or aren't at CMU. Remember, though, that an actual lab is much more interesting than a lobby with pictures, so don't spend too much time there. Also, remember that you will have a lot of time to talk about general things while you are walking from stop to stop, so don't feel that you have to cover all of the general information while you are at the beginning of your tour.

You are always better off having someone from the lab explain the work rather than explaining it yourself. So, when you alert the lab that you are coming, ask them (nicely!) if someone might be available to explain what's going on. If you answer questions rather than lecture, people stay interested longer. Don't be afraid to say "I don't know," but always offer to find out the answer. If people stop being interested, leave that tour stop. Remember that there are other tour stops, and you don't want to have to cut an interesting stop short because you spent too much time at a stop that nobody cared about. Also remember that they may be interested in some stops more than others, so don't feel like you have to spend the same amount of time at each place.

Keep special factors in mind: old people can't walk much, and can't stand for very long. Young people get antsy and need to move from place to place or be doing something or they lose interest.

Pronouncing your words, particularly if you have an accent, is very important. A native English speaker can lose his or her entire audience if they mumble, swallow their words, or talk too quietly. Speak up; there are a lot of background noises, and in a big tour group, nobody out of arm's reach will hear you if you don't shout. Make sure that people in the back can hear and understand you. Dress nicely; for some reason, the better you are dressed, the more people listen to you.

Use raw facts to support anecdotes, not anecdotes to support facts; people like stories rather than lists. Tell engineers stories about how hard the problem is, tell corporate people about applications of the technology, and tell lay people stories about cute things that the robot did. If you know people in the lab, say hi, since everybody feels better if they know that we're all friends here. Remember that you are probably bothering anybody who working in the lab, so acknowledge them when you go in and thank them when you leave.

In my experience, there are four basic types of tours that you might give:

Common Folk

The vast majority of tours will be to academics and other engineering folk that are interested in the same kinds of details that you are: how things work, what the hard parts or the problem were, possible applications, etc. If the group is talkative, you'll save some energy by just giving the high-level overview, and maybe a video, and letting their questions cover the rest. If they're tired, mute, or boring, then you might have to drag them through the details by their teeth. If this gets arduous, simply leave the tour stops sooner.

V.I.P.

A few tours will be for venerable or wealthy people, and they will usually be interested in a very narrow range of things. They are mostly interested in high-level details, but are usually engineers, so treat them as such. Sometimes they've been funding the engineering rather than doing the engineering for quite a while, so be aware of their technical expertise before you barrage them with details about which they don't care.

Prospective Students

A few times a year, we get prospective undergraduate and graduate students that want a tour of the facilities. For obvious reasons, these are more common around the open house, but they happen sporadically throughout the year. On these tours, the specific research that you present is not as important as the picture of the Institute that you paint. Tell them about the research in the context of what role they might have in that research. Tell them about your own experiences, why we're the best danged robotics program in the country, and be sure to give them contacts for further information if they ask, or even if they look like they're thinking about asking.

Public Relations

We discourage this kind of tour, but once in a blue moon we agree to give tours for a group of people that are non-technical in nature, often to high school students or people from the community. These tours are almost entirely different from any other tour you might give. Remember that these people are not engineers. Phrases like "optimization", "stabilization", and "modified Kalman filter" don't mean anything to these people. Hit only the highest level of details, or you'll lose them. Beware answering highly detailed questions of one member of the tour, since you can bore the rest of the tour to tears. You can sometimes deal with those questions by telling that person to talk to you after the tour. In order of preference, show them: things that move, things that they can touch, things that you can touch, and videos. Stick to cute anecdotes rather than scientific minutiae. These are often the hardest tours to give, so I've tried to mention good things for these tours at the various tour stops.

---

Tour Locations

In the previous editions of this guide, I simply referenced Henry's notes because they are so good. For the sake of consolidation, and because I'm replacing a few of his sections, I've incorporated his notes into the guide. This section is the one that will get out of date the fastest. I will do everything that I can to keep it up to date, but if you notice that something has permanently changed to make these notes inaccurate, please let me know.

Public Relations Display, Smith Lobby

Location and Contacts	Videos	Props	Posters
Smith, with the big, honkin' color copier	None	None	Whatever is on the Public Relations Board (see below)

This is an excellent place to start a tour. If your tour is focussing primarily on field and planetary robotics, you might consider the FRC lobby, instead. This display has information about projects from the entire Institute. The display in FRC has only images of FRC projects. While it's impossible to know what photos will be up at what time, the photos have been fairly stable for a while, so I've made a list here with identifying descriptions. For descriptions of these robots, see the Other Robots and Systems part of this guide.

Above the copier:

• Navlab 1 - 5: three large vans and two cars.
• Navlab 6 - 10: several busses and a few cars.
• Autonomous Helicopter: Hmmm.
• Nomad: Big white robot with large silver wheels, pictured in either the desert or the arctic
• Dante I: Big robotic spider on a frozen volcano slope
• Houdini: Yellow with a plow on the front
• Autonomous Loading System: Dump truck and a front loader
• Demeter: Big harvester in a field
• BOA: Cylindrical robot wrapped around a pipe
• Hip Replacement: A bone, a blue sheet, and a metal probe
• Virtualized Reality: A guy sitting in a dome
• Knowledge-Guided Deformable Registration: A bunch of brain scans
• Informedia: Strips of video stills
• MEMS: Two pictures of chip-looking things

Next to the door:

• Microdynamic Systems Lab: small manufacturing apparatus
• Tesselator: Looks like two white boxes, four wheels, and an arm on top
• Xavier / Amelia: These guys have their names tattooed on them
• Workhorse: Big metal robot

Public Relations Display, FRC Lobby

Location and Contacts	Videos	Props	Posters
FRC, across from the High Bay	None.	False moon terrain	Whatever is on the Public Relations Board

This is also a good place to start a tour, but it is really hard to get to, so you probably won't be able to. It has an enormous poster board that you can point to while explaining the history of the RI, and older robots that we've done. This board only includes FRC robots, however. Ending a tour here also works really well, though you will often have to bring the tour back to where you started. While it's impossible to know what photos will be up at what time, the photos have been fairly stable for a while, so I've made a list here with identifying descriptions. For descriptions of these robots, see the Other Robots and Systems part of this guide.

• Navlab I: Big blue van
• Navlab II: Big military van
• US First Plaque: looks like a class photo. See US First, under Education.
• Workhorse (two photos): Big metal robot
• Artisan: A bunch of X-windows full of graphics
• Lunar Rover Initiative: Robot on the moon
• Nomad (two photos): Big white robot with large silver wheels, pictured in either the desert or the arctic
• RoboCon: Water color drawing of a large console with many screens
• BOA: Cylindrical robot wrapped around a pipe
• Houdini: Yellow with a plow on the front
• T.M.R. Award: Plaque with a picture of a reactor on it
• Ambler: Big red, six-legged robot
• Dante I: Big robotic spider on a frozen volcano slope (black and purple)
• Dante II: The aluminum colored big robotic spider

Medical Robotics Laboratory, Smith 130

Location and Contacts	Videos	Props	Posters
Smith, across from Rainbow Rich LaBarca Costa Nikou	3-D Image overlay 10/97 Six Degree of Freedom Sensing for Hand Surgery and Computer Interface 5/25/99	Hip bone and instrument with clamp for registration Torso to show idea of hip replacement Fish to show that roboticists are people, too Bones and bodies all over the room	HipNav Shape-based ("Pinless") Registration Surgical Education Image Overlay

The notes for this lab were provided by Rich LaBarca in September of 1999. As a special note for this lab, DO NOT MAKE THIS A TOUR STOP WITHOUT FIRST CONTACTING THE LAB. They have many tours and people that they bring through themselves, and it is likely that you will conflict with one of them if you don't warn the people in the lab ahead of time. To operate the videos in this lab, make sure that the TV and VCR in the back right corner are plugged in, and that the TV is in VTR mode.

The central goal of this lab is to develop computer assistive and robotic technologies to meet specific clinical needs.

HipNav
The HipNav system is a collaboration between CMU and the Center for Orthopedic Research at Shadyside Hospital. It is actively used on select patients undergoing a total hip replacement surgery by Dr. Anthony DiGioia (an orthopedic surgeon/civil engineer, and one of the heads of the MRCAS (Medical Robotics and Computer-Assisted Surgery) lab). The number of cases increases regularly, so the best thing to do is ask someone in the lab how many have been done (80+ and counting!).

Key notes on HipNav: The regular surgical tools are not replaced, but AUGMENTED. The HipNav system is PASSIVE, and does not use robots, but robotic technology to help guide the surgeon to his or her preoperative plan. If there are questions about the RoboDoc system, the answer is that it's still in use in Europe, but that system takes control away from the surgeon, which we don't want to do.

HipNav has 2 major parts:

Preoperative planning

Done in a virtual environment currently based on CT images (possibly other modalities). Allows a surgeon to pick both femoral and acetabular components based on CAD data, place them relative to the patient's anatomy and check the resulting range of motion of the leg. The goal is to size and place the components so that the optimal range of motion can be achieved and dislocation of the femur is least likely to occur.

Intraoperative guidance

Takes the preoperative parameters generated by the surgeon, and allows them to be realized in surgery. Uses a tracking camera that detects LED rigid-bodies to localize the position of the pelvis and femur. The surface models are registered to their intraoperative locations using a surface-based registration algorithm on a series of points collected along the intraoperative structures by a tracked probe. The system then provides a series of simple interfaces to assist the surgeon in placing the various components.

Image Overlay
The same basic tracking and registration technology is used with a unique display mechanism to create the illusion of seeing virtual structures (based on CT or other data) inside the actual patient on the operating table. Essentially X-Ray vision.

A display with a half-silvered mirror is tracked and updated in real-time. The half-silvered mirror reflects video images, while allowing the surgeon to see what's underneath at the same time. The desired patient structures, as well as the surgeon's head, are tracked in the same way as HipNav.

Using all this information, a video image can be rendered that accurately represents the hidden structures in the surgeon's field of view. Artificial guides or visual plans can also be added to the scene - registered to their respective structures.

This technology can be used for minimally-invasive surgery, fracture reduction, and any other procedure that requires enhanced visualization.

Adaptive tremor canceling
This work is being done by Cameron Riviere. The description was taken from his website:

There are two types of tremor: physiological tremor, which is present in all human motion, and pathological tremor, caused by injury or diseases such as essential tremor, Parkinson's disease, and multiple sclerosis. Pathological tremor greatly degrades manual control of motion. Physiological tremor causes imprecision in fine motor tasks such as microsurgery.

I developed a new adaptive filter to perform adaptive noise canceling of tremor in human-machine interfaces. I demonstrated its use in three practical applications:

• Online canceling of pathological tremor during computer input via mouse, digitizing tablet, etc.
• Off-line non-stationary quantification of pathological tremor recordings for diagnostic and clinical use
Active control of physiological tremor for use in a handheld microsurgical instrument

Augmentation of manual precision
Again, Cameron Riviere, off his website:

This work may be thought of as a superset of the adaptive tremor canceling work. Tremor is not the only source of manual position error. Hand motion during manipulation contains considerable low-frequency error, or drift. Suppressing this error is difficult because it overlaps in frequency with voluntary motion. Furthermore, little is known about the nature of this component of error, or its origin within the human system.

Because so little is known about I am applying cascade neural networks to learn overall patterns of instrument position error during microsurgery. These neural networks can then be used online to estimate the error in hand motion, and send a drive signal to actuators in the tip of an active handheld microsurgical instrument to compensate for the error in real time.

Virtualized Reality Laboratory, Smith 110

Location and Contacts	Videos	Props	Posters
Smith, across from Suzanne's office Peter Rander Sundar Vedula	Virtualized Reality, 6 min	51 Cameras around room 26 Computers with frame grabbers Big South Park chair to right of entrance	None.

The notes for this lab are copied almost entirely from Henry's notes, though they have been updated since 1997. The tape you need is probably in the bottom VCR. To use it, turn on the monitor, the bottom VCR, put the monitor on Line A, and press play on the VCR. The video shows a sequence of five or six research projects, and is not narrated.

This lab is called the virtualized reality lab. Some of you may be more familiar with the related concept of virtual reality, where a computer is used to interact with an imaginary world as you would in a video game, etc..

Virtualized reality is a slightly different concept. In virtualized reality, a computer is used to interact with a model of a real event e.g. a baseball player swinging a bat, a gymnast doing his/her routine, etc. This involves creating a computer model of the actual event. The traditional method of constructing a computer model is called CAD or computer aided design. In CAD an engineer manually designs the model by drawing every component of every object on the computer. So if he were to model a table, he would have to draw nearly every part individually. While this is a good method of representing the geometry of an object it may not do a good job of representing the general appearance of an object: its color, texture, reflective properties etc. The approach they take in this lab is different.

Originally, they used a method called "modeling by seeing." In this method, many, many video sequences of the same scene are taped and merged into a three-dimensional wire mesh model that captures shape, appearance, and motion. The fifty-one cameras were arranged around a geodesic dome, and the images from each were merged using a method called multibaseline stereo.

Over time, they relaxed the restrictions, and the cameras were allowed to be anywhere in the room, as they are now. The cameras use the black spots on the floor to calibrate, and this allows for a much more flexible arrangement. The number of cameras, while more is always better, is no longer required to be as high as 51. Also, the video has been updated. Information now goes direct to hard drives, which means that while the room used to have 51 VCRs, it now has 51 frame grabbers and 26 computers for video processing. This is a vast improvement over the old arrangement in which someone had to press "record" on 51 VCRs every time they took data.

Once the model of the scene has been created through this process, we can then interact with it using the computer. In this virtualized environment, we can replay the event and view it from any perspective, not just the perspectives of the 51 cameras. They call this capability a soft camera.

Another improvement, which can be seen in the last two segments in the video, is that they no longer create wire-mesh models of objects, but rather use a technique called "video merging." In this technique, the soft camera is created by "merging" several camera views into the desired view. This technique is new, however, and you can see artifacts around the moving objects. The advantage of this technique, however, is that a three-dimensional mesh model need never be created, which vastly reduces the amount of processing overhead necessary.

There are many applications for which such a system could be used. One application is for teaching medical students. For instance, a surgical operation could be recorded on video this way. A trainee could study the actual procedure as recorded by the model and observe the procedure from virtually any view point of his/her chosing. Another application of virtualized reality would be entertainment. Instead of watching a basketball game from the preselected camera views, you could view the game from any perspective you choose.

Microdynamic Systems Laboratory, Smith 112

Location and Contacts	Videos	Props	Posters
Smith, across from Suzanne's office	CNN Minifactories Haptic Interface WYFIWYG Agile Assembly Simulation	Haptic Interface Agile Assembly Platform Factory simulation	None.

These notes are Henry's, with the exception of some of the following text about videos. There are four videos at this laboratory. The CNN video is a very good overview of the lab, and is great for non-technical audiences, but is often missing. The haptic interfaces video is more technical, but is the only video of that work. The two other videos you might show, particularly if the CNN video is missing. One, "WYFIWYG" is "What you feel is what you get," and suggests a new paradigm for user interfaces using the haptic interface. The other, "Agile Assembly Simulation" is a more technical version of the CNN video that is entirely simulation, but explains the primary thrust of the work.

Minifactories / agile assembly

One of the goals of this laboratory is to build better factories. The type of factory we are talking about may not be what you normally think of as a factory. We are not talking about the big steel mills that exist or used to exist in and around Pittsburgh. We are talking about what they call "Minifactories." These are small tabletop factories used to build small electrical and mechanical components, particularly computer parts and consumer electronics. Since these are small items, the factories needed to build them do not need to be very large.

This laboratory is developing some of the technology behind minifactories. One of the areas of research is improving the speed and accuracy of these factories. They will always be a demand to manufacture smaller and smaller items particularly in electronics and computers. They are striving to develop robotic technology that operates at the scale of a micrometer.

Another area of study is making these factories more flexible such that they can be easily modified to manufacture a new product. With most current factories it takes at least 6 months to overhaul an assembly line to manufacture a new product. The goal of this lab is to enable these factories to be reconfigured in a few weeks to manufacture a new product.

(Ask Jay Gowdy to have the simulation of the factory running on one of the workstations. If I remember correctly, use the following commands to view the simulation:
s - zoom
left mouse - move
left and middle mouse - rotate and zoom?
right mouse - menu

Also be sure to point out the prototype minifactory -- the thing that looks like a glorified air hockey table.)

Haptic interface

The goal of this project could be thought of as making a better joystick. They idea is to give computer user's a real sense of touch when they interact with computer. Not only will the user be able to exert forces on the joystick, but the joystick will exert forcing back on the user's hand in response to his/her actions - this concept is called force reflection. The goal is to provide the user with immediate and realistic interaction with a computer generated environment.

Visual/haptic interface to a virtual environment

The goal of this project is to build a virtual world that you can not only see, but also touch. And to create a realistic virtual world where the sensations of touch and vision are combined realistically. This is particularly important if a virtual reality system is used to train someone for a task that combines visual skills with motor skills.

To interact with virtual world the user grabs the knob on the robot end-effector. The computer system then builds a composite image where the user's hand is shown the interacting with an imaginary environment generated by the computer. The apparatus is painted blue so the computer can easily distinguish the user's hand from the background. The system then replaces the actual background with the computer generated background. (Same technology used for TV weathermen)

FRC High Bay, FRC A14

Location and Contacts	Videos	Props	Posters
FRC, best accessible through either the second floor entrance or through the walkway from BOM-D	None	Whatever robots are around	None

There's no telling what will be in the highbay; it's regularly cleared out for a new project or an upcoming demo. For the most part, you will have to ad lib about what you see there. The FRC is in the process of updating all of their PR material, so over the next year, information on projects at FRC will be more available. For now, the RI website and these notes are the best way to go. Most of the robots that you will see there are described in Other Robots and Systems, below. Show people the insides of things when possible, and explain its history, if you know it. Since the High Bay is so large and multi-purpose, there is no contact point, so don't worry about warning people that you are coming. It may, however, be closed if large things are being moved. Just for the sake of completeness, here is a brief list of the systems that were in the high bay in September of 1999:

• Distributed Surveillance & Sensing: the two ATV's
• Navlab IV: The big military vehicle
• RoboCran: RoboCrane and the little red robot
• BOA: Not actually there, but the pipes that used to clean are
• Mars Yard: The little red robot and the big foam "rocks"

Planetary Robotics Building

Location and Contacts	Videos	Props	Posters
Can't miss it	None	Nomad	Nomad / Atacama Nomad / Antarctica Nomad / Lunar Exploration

The only machines to show here are Nomad and the Steel Dragon, plus some of the displays of the Atacama desert trek. All other machines have been removed to have more work space. The main activity going on is the preparation of Nomad for the upcoming antarctic season. The robotic arm that you see will be attached to Nomad, and instrumented so that it can look for meteorites under the ice. In the winter, Nomad will be gone as well, so in the winter, unless you can ad lib really well, this might be a stop you can skip.

The first thing that you might be asked about is the blue paint and the clear Plexiglass enclosure around the Nomad work area. The short answer is: it was supposed to look like blue ice, and it didn't work. This is not a well-known fact, nor one that the PRB people would appreciate us spreading around, but if someone asks, you should know the answer. The Steel Dragon is a robot stuck up in the corner of the PRB, and was last year's entry into the US First competition (see Education). The sand pit, on your right as you walk in, was used to simulate the surface of Mars in early planetary exploration research.

While Nomad is there, there are two projects that you can discuss:

Atacama Desert Trek (1997)
The goal of the Nomad project is to develop robots suitable for planetary exploration. In order to have a robot operate on planetary missions, it must be very reliable and capable of autonomous operation for long periods of time. Currently, they have built the Nomad mobile robot to test the feasibility of robotic planetary exploration. The Nomad robot has been sent to the Atacama Desert in Chile where it covered a 200 km trek over the course of two months. The robot was remotely operated from the Carnegie Science Center in Pittsburgh, NASA AMES Research Center, and from Santiago, Chile during this experiment. During its operation, Nomad sent back live, high resolution images that could be viewed at the Science Center. They let visitors to the Science Center remotely control the robot for short periods of time showing that is possible to have machines "smart" to be driven by novices.

Meteorite Robotics Antarctic Search (1997, 98 and 99)
Carnegie Mellon University in cooperation with NASA, the ANSMET, the Chilean Antarctic Institute, and others will perform a series of robotic and science experiments in Western Antarctica in the upcoming expedition season. The team will validate sensors for identification of meteorites located in a blue ice field in the Patriot Hills area.

In 1998, Nomad will perform autonomous searches for meteorites on ice fields designated by field experts based on high resolution aerial imagery. The robot will demonstrate identification and in-situ characterization of meteorites using learning based classification in addition to the other techniques. Equipped with radar sensors the robot will also search for subsurface meteorites. Once a meteorite is detected, onboard computers will record meteorite position, size, geometry, and depth, and provide the information to the human experts for subsequent retrieval. For this purpose, Nomad is being winterized, and prepared to endure the harsh Antarctic environment.

In the 1999 expedition season a team of two robots will demonstrate cooperative search for meteorites. The robots will search for submerged meteorites and meteorites deposited into areas not covered by human search, e.g. ice fields near crevasses. The robotic search will be under shared control by the expedition team and remote scientists in the continental US. Using a synthesis of surface perception, subsurface sensing, and autonomous searching techniques these robots will become invaluable tools to the ANSMET teams.

Advanced Mechatronics Laboratory, Wean 1334

Location and Contacts	Videos	Props	Posters
Wean, the end of the 1300 corridor, on the right Chris Paredis	Gyrover PR Video RMMS Video (missing) RMMS, Traded and Shared Robotics, Tactile Sensing	DM2 Space Station Arm RMMS MIT/Utah Arm Troikabot Assembly from teaching	None

The Advanced Mechatronics Laboratory does a wide variety of research into innovative robotic systems. Of late, research there has been focussing on manipulators. There are a number of manipulators and groups of manipulators in the lab, and they are described in the following from Henry's notes. Unfortunately, the videos in this room suffer from legs; they always run away, so there's no way of knowing what will be there when you want to give a tour. The RMMS video, mentioned above, is not currently available. The tape labeled RMMS does cover the topic, but doesn't have the best segment on the matter. None of these projects are currently active, but are still technologically relavent.

This lab is also where Ben Brown lives, though he's been doing work on the Toy Initiative and a few other projects as well, lately. On of Ben's most publicized projects is the Gyrover project, described below and in the Gyrover video that has a series of clips of Gyrover's publicity. The best segment is the one from KDKA news, taped at Disney World, which is great for non-technical tours.

RMMS

The primary difference between a robot and a machine is flexibility. Machines are only built to do one thing. In theory robots can be reprogrammed to do many things. However, most robots are still limited by their mechanical configuration. They only have limited reach or workspace. They can only lift specified payloads, and they can only move within a specified accuracy.

The goal of this project was to overcome some of these limitations of a fixed mechanical configuration. This robot is called RMMS. This acronym stands for a Reconfigurable Modular Manipulator System. Unlike most robots, it's mechanical configuration can be easily modified. RMMS consists of a set of interchangeable joint and link units, that come apart and snap together. To build a specific robot, you choose from a stock of these units and select the appropriate characteristics and sizes you want for each unit and then snap all the units together. The software then automatically determines the mechanical configuration of the robot and self-calibrates.

The motivation behind this robot is to achieve greater flexibility at reduced cost. So if a manufacturer decides to revamp his/her product, instead of buying a new robot he can just change the configuration of his existing one, by substituting a few links or joints.

MIT-Utah hand, Troikabot, Adepts, Direct Drive Arm
Much of the work in this part of the lab focuses on developing the dexterity and skill of robots in grasping and manipulating parts. This work began in the mid 1980's with the Direct Drive Arm (purple robot). There is no gearing in the joint motors on this robot. The advantage of not having gearing is that it makes it much easier to analytically model the dynamic behavior of the mechanism. By being able to use an accurate model of the robot dynamics, this robot demonstrated that superior performance could be achieved in terms of accuracy and speed.

Now much of the research focusses on interacting with the environment directly: picking and placing objects, and manipulating objects. One technique they have developed is called visual servoing, where a robot positions itself based on its position relative to various objects in its workspace as measured by the robot's camera. By being able to use vision a robot can accurately position itself and manipulate its environment without being precisely calibrated.

They are also programming robots to mimic human skills in manipulating parts. This technique is called "gesture based programming." This involves building a robotic skill automatically by having the computer system observe a human repeatedly perform the desired task. (The "Human Systems Interface" videotape describes this in more detail).

And finally they are also studying robotic assembly, where this set of robots (troikabot) work together to perform an assembly task.

Chimera and Tactile Robotics
The blue robot, to the right of the door, was used in the early nineties as a platform for tactile sensing and real-time sensor interpretation. As the video shows, the tactile sensor at the end of the arm could be used to intelligently drive the robot along a carved groove in the wood.

Gyrover
From the web page, Gyrover is a single-wheel robot that is stabilized and steered by means of an internal, mechanical gyroscope. Gyrover can stand and turn in place, move deliberately at low speed, climb moderate grades (about 20%), and move stably on rough terrain at high speeds (10 kph or more). It has a relatively large rolling diameter which facilitates motion over rough terrain; a single track and narrow profile for obstacle avoidance; and is completely enclosed for protection from the environment. Gyrover is able to right itself when it falls on its side. With proper design, Gyrover can float and propel itself on water, and may be useful in amphibious operations. The third and latest version is 16" (40 cm) is diameter and weighs 16 pounds (7.3 kg). It carries an onboard computer (486 PC), attitude sensors, videocamera and radio systems for control inputs and video transmission. With appropriate control software, we expect GyroverIII to be able to perform reliable locomotion over a variety of terrains in response to intermittent, high-level commands from a remote operator.

Robot Learning Laboratory, Wean 5310

Location and Contacts	Videos	Props	Posters
Wean, in the middle of the 5300 corridor, on the right Greg Armstrong	None	Robin Marion Little John Amelia Flo Xavier Bonnie Jeeves	None

From their web page,

Our research is driven by the vision that machine learning will soon play a pervasive role in robotics, and in similar application domains of embedded systems. For robots to operate robustly in complex and dynamic environments, they must be able to adapt to changes therein, and continuously improve their performance based on experience. Learning plays also an increasingly important role in the design of such systems, as training, instruction, and trial-and-error learning are often superior to conventional programming.

To turn this vision into reality, we pursue research to find new, more effective ways to make robots learn from experience, and to make machine learning succeed in robotics and beyond. Our research ranges from theoretical considerations and basic algorithmic design to practical implementations and demonstrations. The lab has produced robots as visible as Rhino and Minerva, two mobile robots that went into museums as tour-guides; Xavier, arguably the first autonomous mobile robot on the Web; and Jeeves, which won first place award at the 1996 AAAI mobile robot competition. Our latest robots are LittleJohn and Florence. Some of these robots were developed in collaboration with the University of Bonn, Germany.

For descriptions of the different projects, see the relavent parts of Other Robots and Systems. For identification purposes,

• Robin: Little Pioneer with red bands
• Marion: Little Pioneer with blue bands
• Little John: All-terrain Pioneer
• Amelia: Big and black and says "Amelia"
• Flo: Tall, open, Scout base, and has a face
• Xavier: Big and silver and says "Xavier"
• Bonnie: In the corner to the left of the door, on the desk
• Jeeves: Green cover on a Pioneer

---

Hey, What's That?

This section is just FYI, in case somebody asks "hey, what's over there?" These are locations that either have no labs or have no labs that have something to show on a tour. There's no particular reason to mention these if they don't come up.

Smith, Second Floor

The home of VASC, most of our administrative offices, the VASC computer cluster, and a large number of faculty and graduate student offices. Takeo lives here.

FRC, Second Floor

Offices and computer facilities for FRC faculty, staff, and graduate students.

PRB, Second Floor

It's really a loft, but it contains computers and an office(?) for FRC students and staff.

BOM-D, Second Floor

This is really the ground floor. In terms of the RI, it only contains graduate student offices and the Robotics Lounge. Also on this floor is the CMU credit union, IT (instructional technology, the people who provide TVs, VCRs, and other equipment for classes and lectures), and most of the student offices for the HCII.

BOM-D, Third Floor

Well, it did have a bunch of labs and offices, but now it's just a demolished, off-limits area where construction workers go

Wean, Second Floor

The Calibrated Imaging Laboratory is up there, but has been defunct for a while. Steve Shafer (of "Memorial Pool Table" fame) used to run the lab, but Steve Seitz (a recent addition to the faculty) has recently begun work there. Ben Brown also has his office there.

Wean, Fifth Floor

The 5300 corridor houses an odd conglomeration of RI faculty, students, and labs. Matt Mason's office is up there, but he's been there since before the flood (back when Marc Raibert was doing legged robotics, it was across the hall from Matt). Sebasian Thrun and the Robot Learning Laboratory is up there, and is tightly tied to the CS department, also housed in Wean. Finally, people in the speech group live there, and the speech labs are there too. The Robot Learning Lab is there, described above. Oh, and Raj Reddy's office is at the end of the hall.

This section has information on locations, both on and off campus, that are too far off of the beaten path to include in a tour. You might mention these to make conversation when traveling from one place to another, or if your tour has a particular interest that overlaps with the work being done at one of these locations.

REC

The Robotics Engineering Consortium is an entrepreneurial entity dedicated to the development of products incorporating advanced robotics technologies from Carnegie Mellon University, NASA, industrial partners and other sources. Industrial partners join the REC with the goal of using high technology to gain a greater market share, develop new niche markets, or create entirely new markets within their area of competition.

The REC manages product development as a business. Scientists and professional business people work together to develop full business plans to guide product development toward the goals of the industrial sponsor. Project teams offer a full range of expertise, from design through marketing. The REC patents new concepts, provides entrepreneurial training, and streamlines licensing arrangements.

Our success is measured by the quantity of industrial matching funds, number of products to market, and the length of our development cycles. Other vital indicators include cost effectiveness of next generation components, commonality of architecture and components, and manufacturability of products.

The Center for the Neural Basis of Cognition (CNBC) is a joint project of Carnegie Mellon University and the University of Pittsburgh. The Center was initiated by a major gift from the R. K. Mellon Foundation, and builds on several existing collaborations between the two Universities, including the NIMH Center for the Neuroscience of Mental Disorders, the NSF Research Training Program in Neural Processes in Cognition (which is now the CNBC Graduate Training Program), the Pittsburgh NMR Center for Biomedical Research, and the Pittsburgh Supercomputing Center.

Created in 1994, the CNBC is dedicated to the study of the neural basis of cognitive processes, including learning and memory, language and thought, perception, attention, and planning. Studies of the neural basis of normal adult cognition, cognitive development, and disorders of cognition all fall within the purview of the CNBC. In addition, the CNBC promotes the application of the results of the study of the neural basis of cognition to artificial intelligence, technology, and medicine. The CNBC will synthesize the disciplines of basic and clinical neuroscience, cognitive psychology, and computer science, combining neurobiological, behavioral, computational and brain imaging methods.

Many projects combine one or more of the following methodologies: computational modeling, behavioral analysis of normal behavior and effects of brain disorders on behavior, functional neuroimaging, and electrophysiological recording of neuronal activity in behaving animals. Clinical applications include major efforts to understand schizophrenia, Alzheimer's disease, and disorders of language processing due to acquired or developmental disorders.

The CNBC consists of faculty and research scientists whose work relates to the mission stated above. All such faculty have appointments in one or more coordinating departments. These include the Departments of Biological Sciences, Computer Science, Psychology and Robotics at Carnegie Mellon, and the Departments of Mathematics, Neurobiology, Neurology, Neuroscience, Psychiatry and Psychology at the University of Pittsburgh. This list is likely to grow as the CNBC develops.

Porter Hall
Needs to be written.

Doherty Hall
Needs to be written.

Hamburg Hall
Needs to be written.

Current CMU Robots and Systems

Xavier

A robot, commanded over the web, that navigates corridors and tells bad jokes. This is research? It is (well, except for the knock-knock jokes), when the goal of the research is to develop autonomous mobile robots that can operate reliably over extended periods of time, and can learn and adapt to their environments.

These are the goals of the Xavier project, sponsored by ARPA and the Robotics Institute of CMU. Xavier is built on top of a 24 inch diameter, four-wheeled, synchro-drive base, built by Real-World Interfaces. Sensors include bump panels, a Denning sonar ring, a Nomadics laser light striper, and a color camera mounted on Directed Perception pan/tilt head. On-board computation consists of two 66 MHz Intel 486 computers and an on-board color 486 laptop, all connected to each other via Ethernet and connected to the outside world via a Wavelan wireless card. Xavier runs a distributed, concurrent software system under the Linux operating system.

Amelia

Amelia has substantial engineering improvements over Xavier. It has a top speed of 32 inches per second, while improved integral dead-reckoning insures extremely accurate drive and position controls.

Crown Inspection Mobile Platform (CIMP)

We set out to demonstrate that a robot could generate data, first and foremost video data whose quality inspectors would gladly accept for routine visual inspection, and to deliver the data to an "inspector's workstation" off the airplane.

To allow us to concentrate on inspection data and not inspection equipment transportation, we designed an interim robot whose mobility is limited to the fuselage crown and provided CIMP with a 3D-stereoscopic video system that gives the inspectors remote binocular inspection capability.

CIMP has been successfully operated on a 747 at Northwest (as shown in the accompanying figures) and on a DC-9 at US Airlines. Working aircraft inspectors have been uniformly enthusiastic about the quality and utility of the imagery that the CIMP remote 3D-stereoscopic video system delivers.

Lunar Rover Initiative (LRD)

Carnegie Mellon University, in conjunction with LunaCorp, Inc., of Arlington, Virginia, will conduct the first private lunar mission, landing a pair of teleoperated robotic vehicles on the Moon's surface. A central goal of the mission is to provide the public its first opportunity to directly participate in space exploration. Researcher's at the Robotics Institute at CMU in Pittsburgh, will design and build the Moon rovers. The customers for the mission include a theme park, television network, commercial sponsors, and scientists. A variety of educational events and activities will be coordinated with the sponsors and other interested organizations.

Bookstore Project

The goal is to produce a robot wheelchair capable of navigating Carnegie Mellon's campus, traveling from my office to the Campus Bookstore to fetch a book autonomously. To this end, this project encompasses challenges in vision, navigation, learning, obstacle avoidance in a dynamic world and planning with incomplete information. The project uses a robot chassis that is actually an electric wheelchair! Localization and sidewalk-following will be performed exclusively using passive vision. For an informal discussion of vision and navigation, see the Monologue on Navigation.

Gyrover

Gyrover is a single-wheel robot that is stabilized and steered by means of an internal, mechanical gyroscope. Gyrover can stand and turn in place, move deliberately at low speed, climb moderate grades, and move stably on rough terrain at high speeds. It has a relatively large rolling diameter which facilitates motion over rough terrain; a single track and narrow profile for obstacle avoidance; and is completely enclosed for protection from the environment.

Millibots

Millibots are small semi-autonomous and autonomous robots to be deployed by a larger robot or field agent. Current Millibot modules include processing units, motor controllers, sensors, pan/tilt platforms, RF link transceivers. A common serial protocol is planned for inter-modular communications where actuation, sensing and communication processes will run in a distributed fashion.

Vacationing CMU Robots and Systems
Robots that, for one reason or another, are not at CMU right now, but are still active research projects by CMU researchers.

Rhino

The "Deutsches Museum Bonn" (German Museum Bonn) is the first museum for contemporary technology in Germany. This year, the "Deutsches Museum Bonn" contributes for the first time to the "Museumsmeilenfest" of the city of Bonn. The "Museumsmeilenfest", which takes place May 29 to June 1, is a great cultural event, at which the major museums in Bonn offer special exhibitions and tours. It is surrounded by many cultural attractions in the city of Bonn.

Visitors of the "Deutsches Museum Bonn" will have the opportunity to be shown through the museum by a MOBILE ROBOT. During the "Museumsmeilenfest" visitors can directly participate in several tours, guided by the mobile robot "RHINO". RHINO will explain what the people see, and upon request provide in-depth information concerning individual exhibits.

The RHINO-project is a joint project between the "Institut für Informatik III" of the Rheinische Friedrich-Wilhelms-Universität Bonn and the Carnegie Mellon University (USA). The central scientific goal of the RHINO project is the analysis and synthesis of computer software that can learn from experience. The team believes that an essential aspect of future computer software will be the ability to flexibly adapt to changes, without human intervention. The ability to learn will soon enable robots like RHINO to perform complex tasks, such as transportation and delivery, tours through buildings, cleaning, inspection, and maintenance.

While giving tours in the Deutsches Museum, RHINO can be observed and even tele-operated through the Internet. RHINO will make available on-line camera images recorded in the museum.

Nomad

See the Planetary Robotics Building, above.

Sage

Sage is a permanent addition of the Carnegie Museum of Natural History's Dinosaur Hall exhibit area. Sage is a completely autonomous mobile multimedia exhibit built on top of the XR4000 robot base by Nomadic Technologies, Inc. It wanders Dinosaur Hall on a planned path and provides video and audio enhancements to the exhibits for museum visitors.

Sage navigates using a single color video camera. Artificial landmarks placed in Dinosaur Hall help it orient during its journeys. Sage also avoids all forms of collisions, using 48 sonar sensors, infrared sensors and tactile sensors covering the bottom half of the robot.

Minerva

Minerva is a talking robot designed to accommodate people in public spaces. She perceives her environment through her sensors (cameras, laser range finders, ultrasonic sensors), and decides what to do using her computers. Minerva actively approaches people, offers tours, and then leads them from exhibit to exhibit.

The goal of the Minerva project is to bring robots closer to people. Recent progress in robotics and artificial intelligence has made it possible to build interactive mobile robots that operate highly reliably in crowded environments. In the next decade, robots like Minerva are expected to become part of many people's lives, where they will assist them in their everyday activities, perform janitorial services, or simply entertain them. This project is carried out jointly by Carnegie Mellon University's Robot Learning Laboratory and the University of Bonn's Computer Science Department III, and sponsored by the Lemelson Center at the National Museum of American History.

Hip Navigation System (HipNav)

The Hip Navigation or HipNav system is being developed jointly by Shadyside Hospital and Carnegie Mellon University to help reduce the risk of dislocation after total hip replacement surgery. The system allows a surgeon to determine the optimal, patient-specific location for an acetabular implant (socket portion of a hip implant), and guides the surgeon to achieve the desired placement during surgery.

Image Overlay

Image overlay, a visualization method, combines 3D computer generated images with the user's view of the real world. In contrast with other image overlay systems, this system provides the observer with an unimpeded view of the actual environment, enhanced with 3D stereo images. The system has the ability to track changes in the observer's view point and transform the computer images to appear in the appropriate location.

Intelligent Microsurgical Instruments

Positioning error is inherent in normal human hand motion. This includes components such as physiological tremor, jerk, and low-frequency wander. For a surgeon performing microsurgery, involuntary hand motion limits the accuracy with which he or she operates. This problem is especially significant in the fields of ophthalmological and neurological surgery. To deal with this problem, we are developing an intelligent active hand-held instrument for ophthalmological microsurgery. This instrument senses its own motion, distinguishes between desired and undesired motion using advanced filtering techniques, and actively compensates for undesired motion by an equal but opposite deflection of its own tip.

Knowledge-Guided Deformable Registration

The goal of this research is to match corresponding anatomical structures across individuals, and to detect possible pathologies. The current image data is Magnetic Resonance Imaging (MRI) of human brains. MRI datasets are volumetric images which provide 3-D anatomical information. They consist of parallel cross-sections scanned along one of three principal axes. The current approach is to deform a hand-segmented and labelled atlas (Courtesy of Harvard Medical School/Brigham and Women's Hospital) to match a patient's brain, so as to segment and label the patient's anatomical structures using information derived from the atlas. The algorithm applies a hierarchy of deformable models to the atlas to match with the patient at increasing accuracy. A prototype, ADORE (Anomaly Detection thrOugh REgistration), is developed to employ the registration algorithm to detect pathologies that cause morphological changes in the brain.

Autonomous Loading System (ALS)

The ALS project is developing technologies with the intent to improve the productivity and reduce the cost of mass excavation. In a typical mass excavation application, a loading machine digs material from a face and dumps the material in a truck, with a throughput of hundreds of trucks per day. The process is repetitive and continues day and night and in most weather conditions. The type of loading machine and truck can vary, as can the material to be loaded which may range from sand to hard rock. Other issues include maintaining safety of all machines and personnel in the loading area and coordination of multiple machines to achieve a prescribed goal.

Demeter

The Demeter project is developing a next-generation self-propelled hay harvester for agricultural operations. Demeter has proven its ability to exceed typical manned productivity and quality, most recently, achieving the project's 100-acre autonomous harvest milestone.

Unmanned Ground Vehicles

We are developing autonomous navigation capabilities for mobile robots driving in complex, unstructured outdoor terrain. Ultimately, the goal of this work is for teams of robots to be able to drive fully autonomously over long distances, i.e., many miles, in unknown terrain. This project is part of DoD Demo III program. The target mobile robot platform for this project is designed by the Demo III prime integrator, Robotic Systems Technology (RST). The technology developed at CMU was also demonstrated using retrofit HMMWVs as part of the Navlab project.

Video Surveillance and Monitoring (VSAM)

The Video Surveillance and Monitoring (VSAM) project is developing automated video understanding technology for use in future urban and battlefield surveillance applications, where human visual monitoring is too costly, too dangerous, or otherwise impractical. Novel image understanding technologies developed under the VSAM project will enable a single human operator to monitor activities over a large, complex area using a distributed network of video sensors. Sample applications include building and parking lot security, monitoring restricted access areas in warehouses and airports, scanning urban battlezones for sniper activity, and performing reconnaissance on the battlefield. The VSAM project is being sponsored by the Defense Advanced Research Projects Agency, Information Systems Office (DARPA ISO), as part of the Image Understanding for Battlefield Awareness effort.

Autonomous Helicopter (HELI)

To develop a vision-guided robot helicopter which can autonomously carry out functions applicable to search and rescue, surveillance, law enforcement, inspection, mapping, and aerial cinematography, in any weather conditions and using only on-board intelligence and computing power.

Bow Leg Hopper

The bow leg hopper is a novel locomotor design with a highly resilient leg that resembles an archer's bow. During flight, a "thrust" actuator adds elastic energy to the leg, which is automatically released during stance to control hopping height. Lateral motion is controlled by directing the leg angle at touchdown, which determines the angle of takeoff or reflection. The leg pivots freely on a hip bearing, and is automatically decoupled from the leg-angle positioner during stance to preclude hip torques that would disturb body attitude. Upright attitude is maintained without active control by allowing the body to "hang" from the hip joint. Preliminary experiments with a planar prototype have demonstrated impressive performance (hopping heights of 50 cm or more), high efficiency (recovers over 70% of the energy from one hop to the next) and low power requirements (45 minutes of operation on a small battery pack). Current experiments are focused on developing control and planning schemes to enable locomotion over discrete "stepping stones" and obstacles.

Informedia Digital Library

The Informedia Digital Video Library uses intelligent, automatic mechanisms that provide full-content search and retrieval from an extremely large (scaling to several thousand hours) on-line digital video library. We will build the library initially using 1,000 hours of video from archives of WQED Pittsburgh, Fairfax Co. VA School's Electronic Field Trips, and the British Open University's BBC-produced video courses.

Established by Carnegie Mellon University and QED Communications (WQED Pittsburgh), the Informedia project integrates regional and national resources and focuses on creating and applying interactive multimedia to learning and communication in grades K-12 and beyond.

Magnetic Levitation Haptic Interfaces

This project advances knowledge about how to give computer users convincingly real haptic (sense of touch) interaction with computers. While there has been some progress in this area, chiefly through the use of back-driven robotic-like manipulators, this is a substantially new approach which promises a qualitative leap in improvement of such capabilities: A user interacts with the computer by grasping a rigid tool whose behavioral description is computed, employing this tool to interact with computed environments which are semantically meaningful in terms of the application. At the same time, the environment exerts realistic forces and torques on the tool's handle which are felt by the user. The vision is one of providing the computer user immediate, high-fidelity, convincingly real interaction with computed environments.

Skinnerbots

We are developing computational theories of operant conditioning. While classical (Pavlovian) conditioning has a well-developed theory, implemented in the Rescorla-Wagner model and its descendants (work by Sutton & Barto, Grossberg, Klopf, Gallistel, and others), there is at present no comprehensive theory of operant conditioning.

Uranus

Uranus is a mobile robot used for developing 3D mapping and sensing. The mobile base provides 3DOF motion by utilizing a novel set of wheels that can move the robot along any path and any orientation along that path. Each wheel is individually controlled by a brushless DC servo motor. The chassis has two levels; the lowest has mechanics and onboard lead-acid gel cell batteries, and the upper level houses computing and electronics. Above the second level is a baseplate with a regular grid of threaded holes to provide for placement of many sensors, additional computing etc for specific experiments.

Uranus has undergone several upgrades to feedback, motion controllers, computing, power systems, and mechanism over the years. The current system uses resolvers for feedback and commutation at each wheel, onboard small motor controllers, linked to a VME cage with motion control card, computing, and net access.

Uranus is still in use and primarily uses a multi-baseline stereo set-up.

Retired CMU Robots and Systems

The Workhorse

This was one of the first robots built by the Institute. It was designed for cleaning up contaminated environments -- chemical or radioactive contamination. Several similar CMU robots, the Remote Reconnaissance Vehicle (RRV), the Remote Core Sampler (RCS), and the Remote Work Vehicle (RWV) were sent into 3-mile Island in the middle 80's. These remotely controlled robots were used for general utility work, e.g. knocking down walls, washing contaminated surfaces, transporting materials, etc. As well they were also for collecting radiation samples and radiation mapping.

Jeeves

Needs to be written.

Ambler

The Ambler represents an integrated, self sufficient system that was used to provide NASA mission managers with the confidence that legged vehicles are a realistic alternative to wheeled rovers for lunar and martian exploration. The aluminum Ambler has two sets of stacked legs, with three legs per stack. These legs separately lift, advance and circulate to their original positions, much like an egg beater. The martian explorer prototype was a legged platform. It was big and heavy, but demonstrated the ability of a walking machine on difficult terrain.

Dante II

Dante II, a tethered walking robot developed by the CMU Field Robotics Center (FRC), made news in July 1994, when it descended into Mt. Spurr, a volcano on the Aleutian Range in Alaska. High temperature, fumarole gas samples are prized by volcanic science, yet their sampling poses significant challenge. In 1993, eight volcanologists were killed in two separate events while sampling and monitoring volcanoes. The use of robotic explorers, such as Dante II, opens a new era in field techniques by enabling scientists to remotely conduct research and exploration.

Tessellator

Tessellator inspects and waterproofs each of the 17,000 tiles that coat the space shuttle's underside, saving humans a laborious task that lasts from the time the shuttle lands at Kennedy Space Center until just before liftoff. By inspecting tiles more accurately than the human eye, Tessellator reduces the need for multiple reinspections. It also injects into each tile a toxic waterproofing chemical, which prevents the lightweight, silica tiles from absorbing water. Human workers have had to wear heavy suits and respirators to inject the chemical, all the while maneuvering in a crowded work area.

Rosie

ROSIE was a mobile worksystem for selective equipment removal tasks (SERS), developed at Carnegie Mellon University and RedZone Robotics, Inc. for the Department of Energy, for testing at Oak Ridge National Laboratory, Oak Ridge, TN and deployment within the reactor building of the CP-5 reactor at Argonne National Labs.

RoboCon

The human operation and telerobotic and supervisory control of sophisticated and remote decontamination and decommissioning (D&D) robotic systems is a complex, tiring and non-intuitive activity. Since D&D and selective equipment removal (SER) are going to be a major future activity in DOE's ER&WM cleanup agenda, it seems appropriate to utilize an operator control station and interface which maximizes operator comfort and productivity. Carnegie Mellon University (CMU) proposes to develop a state-of-the-art robot operator control station with standard hardware and software control interfaces to be used on a variety of D&D robotic systems currently under development by the OTD. The purpose of this system is to provide a reconfigurable operator interface platform, applicable across D&D robot systems, allowing for cost-effective testing and deployment of various robot systems for demonstration and field-use purposes.The benefit is to be seen in the ability to control different robot systems through simple interchange of interface modules mounted to the operator's chair, and the porting/development of interface display software to a common computing and programming platform. Cost savings can be realized through this system, since it represents a powerful and re-configurable test platform for evaluating the various robot systems currently available or under development for the OTD D&D, Tanks and Mixed Waste focus groupsprograms. The proposed system consists of a large multi-screen projection-TV system framed on both sides by several high-resolution TV monitors, stereo speakers, a reconfigurable operator console and control chair module with various removable interface modules (such as joysticks, buttons, touch-screen, etc.), all ergonomically mounted on a raised platform and integrated with the display and control electronics. The embedded computing consists of computing racks to operate the consoles and to house the robot-control and interface computing. The console computing consists of a dedicated processor system operating communicating with other hardware and interfaces via NDDS over ethernet, serial or parallel interface.

Neptune II

The Field Robotics Center (FRC) at Carnegie Mellon University developed a mobile robot to inspect interiors of fuel storage tanks, both above-ground (AST) and under-ground (UGST). Neptune inspects tanks for visible deterioration, using a video camera and inspects for wall and floor corrosion, using add-on sensors, as part of its standard capabilities. The system included a self-contained portable unit, which attaches to storage tanks. This unit deploys a small, reconfigurable robot, driven by dual magnetic tracks, carrying a camera and lights, navigation system, ultrasonic testing (UT) sensors, or other sensing probes. The robot's reconfigurable design allowed it to fit through large tank openings (20" O.D.) using a parallel-crawler configuration.

Houdini

The FRC has developed a capable mobile robot to gain access to, and move around inside a tank, deploying capable waste movement and handling tools such as a backhoe and plow, to help extricate the waste from the tank by moving waste to a central waste extrication system.

Dante I

Needs to be written.

BOA

Most of the steam and process-piping in DOE facilities is cladded and insulated with asbestos containing materials (ACMs) which will have to be removed before any decontamination and dismantling (D&D) activity. Due to the carcinogenic nature of asbestos flyings and radiological contamination, and abatement regulations from the EPA and OSHA, manual removal is estimated to be rather costly and lengthy. Current methods require substantial infrastructure in terms of scaffolding, containment areas, and air monitoring, resulting in low levels of removal efficiency. A mechanical removal system, dubbed BOA, is being developed, which can be remotely emplaced and is able to crawl on the outside of different-sized pipe to allow complete removal of lagging and insulation while wetting the ACM and encapsulating the stripped pipe, and bagging the removed insulation in-situ. Careful attention to vacuum and entrapment air flow will ensure that the system is able to operate without a containment area while meeting local and federal fiber-count standards. Current plans are to target process piping ranging in diameter from 4 to 8 inches in OD. The advantages of this system are to be seen in the areas of (i)increased material removal efficiency, (ii) reduction in required abatement personnel, (iii) fully contained and sealed operations, and (iv) removal & packaging for easy processing/disposal.

No Hands Across America! (NHAA)

During this tour of America, which was sponsored by Delco Electronics, AssistWare Technology, and Carnegie Mellon University, two researcher from CMU's Robotics Institute "drove" from Pittsburgh, PA to San Diego, CA using the RALPH computer program. RALPH (Rapidly Adapting Lateral Position Handler) uses video images to determine the location of the road ahead and the appropriate steering direction to keep the vehicle on the road. (The researchers handled the throttle and brake.)

Motion Planning for Serpentine Robots

Needs to be written.

Neptune Mobile Robot

Neptune is a three wheeled robot trike. The front wheel both drives and steers and the two rear wheels are passive. It's a very simple motion control system and even simpler power system. This means it works, and works well, and worked for over six years through different projects.

The steering and drive motors are synchronous AC motors controlled through relays. An onboard 68K board provided timing signals and turned relays on and off for motors and accessories.

Neptune's sensors included stereo vision and a ring of ultrasonic time-of-flight sensors.

Non-CMU Robots of Interest

Sojourner

Needs to be written.

FIDO

Needs to be written.

The One that Crashed on Mars in September of 1999

Needs to be written.

Insect Robots

Needs to be written.

Do-be Do-be To-do

• Run through an HTML verifier
• Need to add links where appropriate.
• Need better way to incorporate Henry's notes
• Needs more information from people who know more than me.
• Needs to be proof read
• Timestamp sections
• Important personel
• Describe false moon terrain
• Alphabetize Other Robots and Systems
• Virtualized Reality Lab video explanation
• HH C108
• Fill out Other Robots and Systems
• Where/what is the MegaLab "Human Systems Interface" video?
• Videos need specific dates and label information
• Make it print nicer
• Add BOM-D A Floor
• Two CALD seminar descriptions
• Fix all Needs to be written's
• Mobile Robot Contest
• Education
• MegaLab: what's the little blue guy
• Dave Hershberger: What's Amelia doing?
• Change FRC, BOM-D to Newell-Simon
• More things for non-technical audiences
• Shorter descriptions of other robots and systems
• Update stats on number of people in RI

• 1st Edition, created, Salvatore Domenick Desiano, 11/17/98
• 2nd Printing, minor corrections, Salvatore Domenick Desiano, 11/18/98
• 2nd Edition, major overhaul, Henry's tour notes incorporated, Salvatore Domenick Desiano, 9/27/99

Return to the RoboOrg homepage.

This page can be found on the world wide web at http://www.ri.cmu.edu/~sal/ri-tours.
Maintained by Salvatore Domenick Desiano (sal@ri.cmu.edu).