Research Summaries

Back In-stride Optimal Motion Planning/re-planning for MCM Missions using Optimization

Fiscal Year 2019
Division Research & Sponsored Programs
Department NPS Naval Research Program
Investigator(s) Kaminer, Isaac I.
Kragelund, Sean P.
Sponsor NPS Naval Research Program (Navy)
Summary Unmanned systems are playing an increasingly important role in naval mine countermeasures (MCM), but these systems have yet to realize their full potential for reducing mine clearance timelines. In fact, mine hunting missions are largely conducted in sequential phases. Often, a single vehicle class or sensor type is used to conduct each phase, and extensive post-mission analysis is required in order to plan and execute subsequent phases. This research seeks to overcome the limitations of the sequential search paradigm by enabling simultaneous deployment of heterogeneous vehicles with complementary capabilities. Specifically, we propose to investigate new algorithms and computational tools for generating optimal search trajectories for heterogeneous, multi-agent teams in a real-time, event-based framework. The proposed research will address three inter-related aspects of this motion planning problem. First, motion plans must consider vehicle-specific dynamics and sensing capabilities for each team member, probabilistic target models, and MCM mission objectives like mine detection probability or area clearance rate (ACR). Second, the planning framework must be capable of responding to detection events so that underlying probabilistic models “evolve” as a function of individual search vehicles’ progress. Finally, this capability requires real-time algorithms for in-stride re-planning that utilizes appropriate vehicles and sensors to classify detected targets as mines or non-mines (false positives).

1. Can real-time planning/re-planning on unmanned vehicles reduce MCM timelines compared to standard, pre-mission planning for sequential missions?
2. Does in-stride re-planning produce comparable risk thresholds in less time?
3. When agents detect a mine target, how should the underlying target distribution evolve?
4. Can this framework address "false positives?"

Deliverables include a student reports, a final report, a manual and well annotated real-time code.
Keywords
Publications Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal
Data Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal