Crowd Navigation in a Multi-Room Environment: a Model Predictive Control Framework for Mobile Robots

Giovanbattista Gravina, Francesco D'Orazio, Michele Cipriano, Tommaso Belvedere, Giuseppe Oriolo

Paper Code

Abstract

Mobile robots operating in human-populated environments must navigate complex, multi-room spaces while ensuring safety, i.e. generating collision-free motion. This paper presents a sensor-based Model Predictive Control (MPC) scheme designed for safe crowd navigation in such non-convex environments. The proposed framework decomposes the free space into a set of overlapping convex regions to construct a topological graph, enabling a high-level planner to compute optimal sequences of traversable areas. To effectively perceive the crowd, the system employs a robust perception pipeline that fuses 2D LiDAR data with semantic information from an RGB-D camera, utilizing Kalman Filters (KFs) to estimate and predict human motion. These predictions are integrated into an MPC controller which generates robot commands by enforcing safety through Discrete-Time Control Barrier Functions (DT-CBFs), ensuring the robot avoids collisions while remaining within navigable regions. The approach is validated through high-fidelity simulations and real-world experiments using the TIAGo mobile manipulator. Results demonstrate that integrating vision-based semantic data with geometric constraints significantly improves collision avoidance and success rates in cluttered, multi-room scenarios.

Architecture

Project Overview Diagram
Block diagram of the proposed framework. The red block processes raw observations \(\mathcal{O}\) to produce crowd motion prediction \(\mathcal{P}\) via detection, estimation and prediction modules. The yellow block decomposes the environment in a set of convex regions and constructs the graph \(\mathcal{G}\). The green module plans the sequence of regions \(\mathcal{S}\) to reach the goal position \(\boldsymbol{p}^{\text{goal}}\) and the blue module generates the control inputs \(\mathbf{u}\) via MPC.

Simulations

The following Gazebo simulations show the TIAGo mobile manipulator navigating various multi-room environments populated with static obstacles and moving humans, while being empowered by the proposed MPC scheme.

Experiments

Clips of the real robot traversing doors and avoiding humans in diverse dynamic environments.

Full Project Video

Watch the following video for a comprehensive overview of the project.