Mutual information measures the dependence of two random variables, such as a noisy sensor measurement in the future and the current best estimate of a target’s location. Mutual information is an ideal metric to maximize for improving search. Further, given constraints in mission duration, the agent must be able to properly decide where to search, and for how long, before returning to base. Scenarios involving noisy sensors or targets that appear or disappear over time require multiple passes to correctly determine whether or not a target is present. There are several reasons why generating a multipass coverage plan is advantageous for target search. However, such environments may be subject to change, requiring quick decision making while acting on a previously generated plan. It would be advantageous to have aerial vehicles autonomously fly at low altitudes navigating over rubble searching for survivors. Ĭonsider a scenario where an earthquake or flood causes a building to collapse, trapping survivors who are difficult to access by ground. Interest in using drones for search and rescue and assessing dangerous situations has also been increasing in local fire fighting departments. In the 2013 earthquake that hit Lushan, China, Chinese researchers provided an autonomous rotary-wing UAV to assist search and rescue efforts. Survival rates of earthquake victims continually drop significantly the longer the delay before rescue, with a sharp drop off at around 48 h. Victims in such disasters have a greater chance of survival the sooner they are discovered conducting an extensive search even just a few hours faster will save many more lives. The proposed algorithm is based on best first branch and bound and is benchmarked against state of the art algorithms adapted to the problem in natural Simplex environments, gathering the most information in the given search time.Įarthquakes and similar disasters in remote areas pose a challenge for relief efforts when the transportation infrastructure is damaged. To the authors’ knowledge this is the first attempt at efficiently solving multipass target search problems of such long duration. The algorithm is capable of generating long duration dynamically feasible multipass coverage plans that maximize mutual information using a variety of techniques such as ϵ-admissible heuristics to speed up the search. We present an anytime algorithm for autonomous multipass target search in natural environments. In addition, planning multipass trajectories requires evaluating path dependent rewards, requiring planning in the space of all previously selected actions, compounding the problem. If unanticipated changes occur in an uncertain environment, new plans must be generated quickly. Further, mission duration constraints must be handled accordingly, requiring consideration of the vehicle’s dynamics to generate feasible trajectories and must plan trajectories spanning the entire mission duration, something which most information gathering algorithms are incapable of doing. Motion planning for multi-target autonomous search requires planning over an area with an imperfect sensor and may require multiple passes, which is hindered by the submodularity property of mutual information. When searching for targets, maximizing mutual information of future sensor observations will minimize expected target location uncertainty by minimizing the entropy of the future estimate. Entropy can be used to quantify the generation and resolution of uncertainty. Often, finding survivors a few hours sooner results in a dramatic increase in saved lives, suggesting the use of drones for expedient rescue operations. Consider a disaster scenario where search and rescue workers must search difficult to access buildings during an earthquake or flood.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |