LABUST has taken part in many scientific research projects relating to underwater systems and technologies in various roles. This includes numerous international (FP7, H2020, ONRG, INTERREG,...) as well as national projects.
Laboratory for Underwater Systems and Technologies (LABUST), CoE MARBLE, and IEEE OES University of Zagreb Student Branch Chapter invite you to a talk "Reputation Management in Open Robot Swarms through Crypto-Economics" held by Francesco Cerri from Piacenza, Italy, as part of the MARBLE project.
The lecture will take place on Tuesday, April 15th, 2025 at 10 a.m. in Banjavčićeva 1a (meeting room 3-21/3rd floor). The lecture will be held in person in English language and is estimated to last 60 minutes, including questions.
A summary of the lecture as well as the lecturer's biography can be found below.
SUMMARY:
A swarm of robots presents a great efficiency in a foraging task, even without the presence of a global or external positioning system: by building a dynamic chain of agents, they can implement a distributed localization system. Yet, this comes at the cost that even a single incompetent or faulty robot could tamper with the operation of the whole swarm.
Moreover, one could envision a task where the swarm is open, and different parties could enroll robots in the swarm: in this case, no assumption on the cooperativeness of the agents can be expressed. Both malfunctioning and malicious robots can be unified in the Byzantine category: information shared by them cannot be considered trustworthy a priori.
Cooperativeness can be fostered through a system of rewards and punishments implemented via an information market, which awards tokens to robots that comply.
This work shows how econometrics analysis can be used by robots to define a distributed reputation system, allowing the identification of Byzantine agents, introducing tolerance in the swarm, and improving its effectiveness.
BIOGRAPHY:
Francesco Cerri is a graduated engineer in Automation from University of Bologna.
Graduated in vehicular engineering at Politecnico di Milano, he shifts his focus towards robotics, with a particular interest in multi-agent optimization, decision making and learning.
His master thesis analyses how a swarm of robots can become tolerant against deceitful information using a blockchain market, even without assumption on agents' cooperativeness.
He also served as project lead for the Alma-X rover team, which scored the fifth place in the European Rover Challenge 2021.
Francesco is interested in emerging behaviour, self-organization, and is fascinated by the interdisciplinary aspect of robotics.