INCOST - Intelligent Construction Site Task Scheduler

Construction and deconstruction sites increasingly rely on robots for transport, manipulation, inspection, and heavy lifting. These machines rarely work alone — they share tight, constantly changing spaces with each other and with human workers, and a robot that navigates or manipulates perfectly on its own can still collide with a crane’s lifting zone or block a colleague’s path. INCOST develops the higher-level coordination that lets a heterogeneous robot team work together safely and efficiently on a real site. The project treats physical site zones as time-dependent, schedulable resources rather than fixed map regions. Building on the AI Domain Definition Language (AIDDL), it combines hierarchical task decomposition with temporal reasoning to produce a coordinated schedule, then derives a Map of Dynamics that records when each zone is occupied or free. Other agents, such as an independent inspection robot, reuse this map to plan conflict-free routes that run concurrently with the main operation. The approach has been demonstrated on a modular steel-beam deconstruction scenario with a heterogeneous robot team, in simulation and on physical hardware.

In Collaboration with: Dr. Uwe Köckermann, Dr. Martin Magnusson (Örebro Universitet)

Duration of the Project

November 2025 until April 2026

Researchers

Pranav Shevkar (shevkar@ip.rwth-aachen.de), Görkem Ertemli (ertemli@ip.rwth-aachen.de)

Project Video

Funding

This project is funded by the euROBIN Horizon Europe project (Grant Agreement No. 101070596) through its 1st Open Call subgrant INCOST (Agreement No. euROBIN_1OC_CP_9)

Project Partners