Joint Perception and Temporal Logic Planning for Distributed Agents in Dynamic Environments

From u-t-autonomous.info
Jump to: navigation, search

Technical point of contact: Behzad Kamgar-Parsi, Office of Naval Research
Period of activity: 2018-2022
Collaborators: Lu Feng (PI, Univ of Virginia), Zsolt Kira (Georgia Tech), Pratap Tokekar (Univ of Maryland)

Overview of the Project

The objective of this project is to develop theory and algorithms for decentralized perception and temporal logic planning for a team of autonomous agents collaboratively operating in dynamic, uncertain and unstructured environments where large volumes of heterogeneous streaming data are collected.

A key challenge is how to derive intelligence from the massive, distributed, and diverse data sources, and enable the rapid decision-making of autonomous agents. We propose a unified theoretical framework of decentralized perception and planning to address these two needs. The project has three research thrusts that are tightly integrated into a rigorous closed-loop framework.

  • Thrust I – Decentralized perception for planning: How to integrate information from multiple sensing modalities and diverse sources, filter out irrelevant information, and learn distributed and shared representations of the environment in a team of agents?
  • Thrust II – Collaborative decentralized temporal logic planning: How to make planning decisions for the distributed team of autonomous agents based on the semantic world model with uncertainty learned in Thrust I, subject to a rich set of temporal logic specifications representing mission goals and problem constraints?
  • Thrust III – Using planning and communication to guide perception: How to close the loop by using information from planning and communication to guide and achieve task-aware, active perception?
caption