Computational Sensors

Takeo Kanade and R. Bajcsy
Report from the DARPA Workshop, May, 1993.

  • Adobe portable document format (pdf) (467KB)
Copyright notice: This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.

This report is a result of a workshop on Computational Sensors that was organized and held at The University of Pennsylvania on May 11-12, 1992. It presents a summary of the state of the art in computational sensors and recommendations for future research programs. Approximately 40 people were invited from academia, government, and industry. The workshop hosted several key presentations and followed them with group discussion and summary sessions.

Traditionally, sensory information processing proceeds in three steps: transducing (detection), read-out (and digitization), and processing (interpretation). Micro-electronics technologies have begun to spawn a new generation of sensors which combine transducing and processing on a single chip - a computational sensor.

A computational sensor may attach analog or digital VLSI processing circuits to each sensing element, exploit unique optical design or geometrical arrangement of elements, or use the physics of the underlying material for computation. Typically, a computational sensor implements a distributed computing model of the sensory data, including the case where the data are sensed or preprocessed elsewhere. Combining computation and signal acquisition into a single chip results often in not only performance improvement but also totally new capabilities that were not previously possible. Finally, the workshop made several important recommendations.

1. Create a research and development program in computational sensors. The program must have the following characteristics:

  • Interdisciplinary - the program must include sensing, algorithms, VLSI, material, and applications;
  • Multi-modal - the program must deal with not only the image or visual modality, but also with other sensing modalities including tactile, acoustic, pressure, acceleration, chemical, and so on;
  • Prototyping-oriented - individual projects under this program must be oriented toward producing working prototype devices or systems;
  • Applications - individual projects must identify potential applications and possible avenues of technology transfer to real world applications.
2. Improve the infrastructure for research and development of computational sensors:
  • Fabrication facilities - MOSIS (or similar facilities) must be expanded to include technologies for optical and mechanical sensor development;
  • Tools - Tools for designing and testing computational sensors can be far more complicated than they are for standard VLSI design. Standardization, and library and tool development are essential;
  • Education - Hands-on experience must be provided to graduate students;
  • Networking and workshops - Researchers in computational sensors, by its nature, are scattered in multiple fields, and mechanisms; workshops and consortiums must be developed to bring them together.

Associated Center(s) / Consortia: Vision and Autonomous Systems Center

Text Reference
Takeo Kanade and R. Bajcsy, "Computational Sensors," Report from the DARPA Workshop, May, 1993.

BibTeX Reference
   author = "Takeo Kanade and R. Bajcsy",
   editor = "T. Kanade and R. Bajcsy",
   title = "Computational Sensors",
   booktitle = "Report from the DARPA Workshop",
   month = "May",
   year = "1993",