Interview with Dr. Paley

In the course of my research for this article, I had the opportunity to speak with Dr. Derek Paley of the University of Maryland College Park, an Associate Professor in the Department of Aerospace Engineering.[3] He is a recent graduate of Princeton University, having received his doctorate in Mechanical and Aerospace Engineering. He is conducting research in nonlinear dynamics and control theory, specifically related to the cooperative control of multiple vehicles. He is primarily involved in two projects with the University, using a workforce of mostly undergraduate students. He has a few graduate students involved in his research, and seemed enthusiastic about the possibility of adding more graduate students in the near future.

One project is the development of a testbed for decentralized, coordinated control of multiple underwater vehicles using onboard, passive sensors to achieve movement patterns that would allow optimal collection of data. The specific movement patterns and details such as inter-vehicle spacing would depend on the data to be collected. The testbed, then, would be used to verify in situ performance of control laws developed using computer simulations. At present, he doesn't see much room for any automated planning research on this project, but, given the current popularity of swarm robotics research, it seems promising for the future.

The other project Dr. Paley is involved with is called the Glider Coordinated Control System (GCCS). It is part of the Adaptive Sampling and Prediction (ASAP) program being developed for the Office of Naval Research at Princeton University. GCCS is implemented as a central control system for a (small) fleet of undersea robotic gliders that can be used to collect data over long periods and large areas to aid efforts such as hydroacoustic surveillance. Currently, the GCCS has little in the way of automated planning or intelligence, but it does incorporate some simple rules like, “don't steer the gliders into shallow water.” He said that, while his interests are mainly in control theory, there is some room for the addition of more sophisticated planning activities so that the system could be left unattended for longer periods of time without suffering from disturbances arising from weather and tides.

Dr. Paley's work may not seem have a strong tie to AI research, but it was interesting that he seemed receptive to the idea of introducing automated planning concepts into the GCCS. This suggests that there may be other opportunities lurking in projects that take pride in their "classical" roots, if one knows who - and how - to ask.

Significant Projects

There have been many projects in AI planning. The approach taken usually depends on the scope of the desired solution. For real-world applications, computing resources and time are often very limited, so the solution must be highly tailored to the application.

NASA's Remote Agent, Autonomous Sciencecraft Experiment, and the Mars rovers Spirit and Opportunity are examples of application-specific automated planning. These missions are particularly significant because they represent a considerable level of trust in the technologies that are deployed, since millions of taxpayer dollars were at stake.

Projects with a more general approach include the General Problem Solver developed by Newell and Simon in 1957 and its successor, Soar. Soar is currently being developed at the University of Michigan, and claims DARPA among its supporters. Soar has evolved from a planning system to an approach to general AI with planning at its core.

Patrick Doyle's web page on planning provides a long list of specific projects from the last half of the 20th century, and the approaches used by their developers. The Association for the Advancement of Artificial Intelligence (AAAI) web site also contains many web links, both current and historical, on the page covering planning.