Robot Design for Everyone: Computational Tools that Democratize the Design of Robots

Abstract

A grand vision in robotics is that of a future wherein robots are integrated in daily human life just as smart phones and computers are today. Such pervasive integration of robots would require faster design and manufacturing of robots that cater to individual needs. For instance, people would be able to obtain customized smart assistant devices such as a home monitor personalized with their child’s favorite fantasy character, a helping hand robot for their specific art project, or a cleaning robot that can clean hard to access spaces of their homes. However, robots of today take years to be created by experts, and are often not customizable. To enable widespread integration of robots, we wish to democratize the robot design process in order to support rapid creation of custom robots for the people and by the people.

In recent years, advances in digital fabrication technologies and the availability of affordable electronics such as Rasberry Pi, Arduino etc. is enabling rapid creation of smart devices. Unfortunately, currently these technologies are only accessible to experts because of the skills and domain knowledge needed to build with them. To change this status quo, we present a suite of easy-to-use computational tools for enabling the design of a broad class of robotic devices that casual users might want to create.

In particular, we develop tools for designing physical structure and task-specific behavior of robotic devices with various form factors and functionalities. The key strengths of our tools include intuitive visual interfaces that can support user-in-the loop interactive design, parameterized domain-specific models and physics-based simulation that can encode complex design aspects, and efficient algorithms that can search high-dimensional design spaces at interactive rates for user-preferred solutions. Our tools also automate tedious design steps allowing users to focus on the creative aspects of the design process. Finally, we show how simulation-based feedback and data-driven design can be used to lower the barriers to entry for casual users. We validate our tools by fabricating various prototypes and by conducting user-studies with novices.

In the past, design tools such as those offered by Adobe creative suite and Autodesk have revolutionized the creation of diverse digital content ranging from images to animation. The tools presented in this thesis take a step towards enabling the same for the domain of robotics.