FnS - Managing Distributed Situation Awareness in a Team of Agents Tasked to Conduct Search Activity

Add to your list(s) Download to your calendar using vCal

• Sagir Yusuf
• Thursday 12 May 2022, 16:00-16:30
• LG23.

If you have a question about this talk, please contact Mirco Giacobbe.

This project describes how distributed agents operating under strict constraints can update a system level Situation Awareness (a process of collecting, understanding, and predicting system information). The constraints are limited energy (battery), computational power, dynamic environment, and role variation. The problem is addressed by developing (i) a novel way of coordinating agents’ information collection process using geometrically derived path generation algorithms and (ii) a better information presentation technique using Bayesian Belief Network (BBN). Algorithms and methods were developed to describe an effective BBN update process (in terms of both parametric and structure features), BBN training to support situations prediction and uncertainty handling. The evaluation process was described using a team of Unmanned Aerial vehicles (UAVs), either simulated or physical tasked to search forest fires. Results show better performance over the existing methods in terms of agents’ coordination (compared against popular fixed-pattern and pseudorandom approaches) and SA management (compared against propositional networks, ontology, and concept maps). Finally, the developed agents’ coordination algorithms demonstrate easy application on real UAVs (drones) as tested on a team of UAVs tasked to search one of the Birmingham’s parks. So, even if you are not interested in the algorithms and methods, at least you will learn how to fly and coordinate multiple drones for your summer fun time.

This talk is part of the Facts and Snacks series.

This talk is included in these lists:

• Facts and Snacks
• LG23

Note that ex-directory lists are not shown.