Carnegie Mellon University THE ROBOTICS INSTITUTE
Thesis ProposalJonathan Michael Butzke
Wednesday, December 2, 2015 GHC 810212:00 p.m.
Maxim Likhachev Chair
Sebastian Scherer
Nathan Michael
Dan Lee University of Pennsylvania
Thesis Committee
Planning for a Small Team of Heterogeneous Robots:from Collaborative Exploration to Collaborative Localization
Abstract Robots have become increasingly adept at performing a wide variety of tasks in the world. However, many of these tasks can benefit tremendously from having more than a single robot simultaneously working on the problem. Mul>ple robots can aid in a search and rescue mission each scou>ng a subsec>on of the en>re area in order to cover it quicker than a single robot could. Alterna>vely, robots with different abili>es can collaborate in order to achieve goals that individually would be more difficult, if not impossible, to achieve. For example, in an explora>on scenario, a ground robot can carry a large baEery capacity providing it with a higher endurance, and thus could spend longer searching than an aerial vehicle. Conversely, the ground vehicle by its very nature is more limited in the terrain that it can traverse than the aerial vehicle. In order to adequately explore a large, obstacle strewn environment, the two robots will have to intelligently determine when is the best >me and where is the best posi>on to employ each of them. In these cases, mul>-‐robot collabora>on can provide benefits in terms of shortening search >mes, providing a larger mix of sensing, compu>ng, and manipula>on capabili>es, or providing redundancy to the system for communica>ons or mission accomplishment. One principle drawback of mul>-‐robot systems is how to efficiently and effec>vely generate plans that use each of the team members to their fullest extent, par>cularly with a heterogeneous mix of capabili>es.
Towards this goal, I have developed a series of planning algorithms that incorporate this collabora>on into the planning process. Star>ng with systems that use collabora>on in an explora>on task I show teams of heterogeneous ground robots planning to efficiently explore an ini>ally unknown space. These robots share map informa>on and in a centralized fashion determine the best goal loca>on for each taking into account the informa>on gained by other robots as they move. This work is followed up with a similar explora>on scheme but this >me expanded to an air-‐ground robot team opera>ng in a full 3-‐dimensional environment. The extra dimension adds the requirement for the robots to reason about what por>ons of the environment they can sense during the planning process. With an air-‐ground team, there are por>ons of the environment that can only be sensed by one of the two robots and that informa>on informs the algorithm during the planning process. Finally, I extend the air-‐ground robot team to moving beyond merely collabora>vely construc>ng the map to actually using the other robots to provide pose informa>on for the sensor and computa>onally limited team members. By explicitly reasoning about when and where the robots must collaborate during the planning process, this approach can generate trajectories that are not possible if planning occurred on an individual robot basis.