Dr. Lukas Chrpa

Faculty of Mathematics and Physics
Charles University in Prague
chrpa [AT] ktiml.mff.cuni.cz
Assistant Professor
Faculty of Electrical Engineering
Czech Technical University in Prague
chrpaluk [AT] fel.cvut.cz

Passionate about making machines think to improve quality of humans everyday life.

Research Interests

DBLP Profile
Google Scholar Profile

Key facts

Since 02/2017 -- Assistant Professor at Czech Technical University at Artificial Intelligence Center (AIC)
Since 01/2017 -- Researcher at Charles University at Department of Theoretical Computer Science and Mathematical Logic
11/2011 - 01/2017 -- Research Fellow at University of Huddersfield at PARK (Planning, Autonomy and Representation of Knowledge) research group
02/2010 - 10/2011 -- Researcher at Czech Technical University
10/2005 - 09/2009 -- PhD Student at Charles University


Title: MoRePlan: Modelling and Reformulating Planning Problems
Role: co-PI
PI: Roman Bartak
Funder: Czech Science Foundation
Years: 2018 - 2020
Goals: Develop modelling and reformulating techniques addressing "accidental complexity" in planning domain models with emphasis also to more expressive forms of planning (e.g., temporal, conformant) for making a strong impact in real-wold applications of planning.

Title: Balancing Deliberative and Reactive Behavior of Intelligent Agents
Role: PI
Funder: Czech Science Foundation
Years: 2017 - 2019
Goal: Develop a framework for balancing deliberative and reactive behavior of intelligent agents, so they can efficiently acquire their goals even in more dynamic environments.

Title: Machine Learning and Adaptation of Domain Models to Support Real Time Planning in Autonomous Systems
Role: co-I
PI: Lee McCluskey
Funder: EPSRC UK
Years: 2012 - 2016
Goals: The project aimed on designing, modelling and learning planning domain models for Oil-drilling industry. The project was done in collaboration with University of Edinburgh and Schlumberger

Title: Tactical Agentscout
Role: co-I
PI: Michal Pechoucek
Funder: US Army, CERDEC
Years: 2010 - 2011
Goals: The projects aims at investigation of problems and challenges in implementation of adversarial behavior in dynamic environments from game theoretic perspective and investigation of issues in long-term (deliberative) vs. short-term (reactive) multi-agent planning in adversarial scenarios. The empirical evaluation of the project aimed at validation in multi-agent simulations of urban warfare scenarios including heterogeneous teams composed of UAVs, VTOLs and UGVs.

Title: LeCoS: Merging machine learning with constraint satisfaction
Role: co-I
PI: Filip Zelezny
Funder: Czech Science Foundation
Years: 2008 - 2010
Goals: Use Machine Learning and Constraint Satisfaction techniques in "hard" problems (e.g. planning)


Co-organizer of the 8th International Planning Competition (IPC-2014) - Deterministic track.

Co-organizer of the 5th International Competition on Knowledge Engineering for Planning and Scheduling.

Co-organizer of the WIPC - Workshop on the International Planning Competition 2015.

Co-organizer of the Workshop of the UK Planning & Scheduling Special Interest Group (PlanSIG) 2016.

Co-organizer of the Workshop on Knowledge Engineering for Planning and Scheduling (KEPS) 2017.


PTT (Planning Task Transformer) is a toolkit for reformulating planning tasks by learning macro-operators and/or entanglements. In particular: Download sources here
Download examples here