Michael Cashmore is a Research Associate in the Department of Informatics at King's College London. There he is a member of the Planning Group. His research interests are in AI Planning for Robot Autonomy, Trusted Autonomous Systems, Machine Learning and Planning, Costraint Programming, and involves the projects below.


The ROSPlan framework provides a generic method for task planning in a ROS system. ROSPlan encapsulates both planning and dispatch. ROSPlan has a modular design, intended to be modified. It serves as a framework to test new modules with minimal effort. Alternate approaches to state estimation, plan representation, dispatch and execution can be tested without having to write an entire framework.

You can find more details about this on the ROSPlan website:

Hybrid Planning with SMTPlan

SMTPlan+ is a planner for hybrid systems. It supports all the features of PDDL+, including exogenous events and continuous processes, providing an SMT encoding of the PDDL+ models. SMTPlan+ can handle linear domains as well as domains with nonlinear polynomial change.

PDDL+ is the extension of PDDL that allows modelling of mixed discrete-continuous domains, and it follows the Hybrid Automata semantics. Dealing with hybrid systems is becoming more and more an important challenge, as many real-world scenarios feature a mixture of discrete and continuous behaviours.

You can find more details about SMTPlan+ and source code on the project website: