Language Support for Multi Agent Reinforcement Learning: Our paper at ISEC 2020

Our research with TCS Research Labs continues to bear fruit. The Actor based language ESL used for much of our collaborative research now incorporates machine learning capability and delves into Digital Twin technology. We report on some early results at one of our favourite conferences: Innovations in Software Engineering (ISEC) – India’s premier software engineering conference.


Software Engineering must increasingly address the issues of complexity and uncertainty that arise when systems are to be deployed into a dynamic software ecosystem. There is also interest in using digital twins of systems in order to design, adapt and control them when faced with such issues. The use of multi-agent systems in combination with reinforcement learning is an approach that will allow software to intelligently adapt to respond to changes in the environment. This paper proposes a language extension that encapsulates learning-based agents and system building operations and shows how it is implemented in ESL. The paper includes examples of the key features and describes the application of agent-based learning implemented in ESL applied to a real-world supply chain.

The full paper will be available at Middlesex Research Repository at this link. 

The conference and other accepted papers are available here.