An AI-Based Planning Framework for HAPS in a Time-Varying Environment

Jane Jean Kiam, Enrico Scala, Miquel Ramirez Javega, Axel Schulte

PosterID: 2 PDF Poster BibTeX

High-Altitude Pseudo-Satellite (HAPS) is a fixed-wing, completely solar-powered long-endurance Unmanned Aerial Vehicle (UAV). It is a flexible alternative to satellites with fixed-orbits for monitoring ground activities. However, given its light-weight build and weak electro-motors, the platform is rather sensitive to weather and cannot fly around hazardous weather zones swiftly. In this work, we propose a PDDL+ formulation of the mission operation of multiple HAPS, the mission payloads of which could be heterogeneous. To ensure feasibility, the formulation also considers the problem of modeling the platform dynamics, the time-varying operation environment (i.e. wind field, weather-related NoGo-areas etc.), and the heterogeneous tasks to carry out. Additionally, we propose a framework that combines a PDDL+ planner with an Adaptive Large Neighborhood Search (ALNS) approach. The ANLS is developed to guide the task assignment using heuristics different than those used by the automated planner to guide the search of feasible control parameters to navigate the HAPS in time-varying wind field. The task and motion planning are done in an intertwined way within the framework, preserving hence a common decision/search space. Validation tests using a third-party six degrees-of-freedom HAPS simulator and real historical weather data show that computed plans are executable. Benchmarking tests confirm that the implemented framework is beneficial, thanks to its ability to perform task and motion planning in a more tightly-coupled fashion.

Session E5: Robotics & Embedded Applications
Canb 10/27/2020, 17:00 – 18:00
10/31/2020, 00:00 – 01:00
Paris 10/27/2020, 07:00 – 08:00
10/30/2020, 14:00 – 15:00
NYC 10/27/2020, 02:00 – 03:00
10/30/2020, 09:00 – 10:00
LA 10/26/2020, 23:00 – 00:00
10/30/2020, 06:00 – 07:00