Cascading Velocity Modulation for Multi-Agent Path Finding Execution
SeungHyun Park, jae hong Shim, Gyuho Eoh
AI summary
Problem
Real-world multi-robot fleets routinely deviate from planned trajectories due to sensor noise, communication dropouts, and actuator faults, yet existing online execution layers react with aggressive binary stop-and-go commands that create inefficient ripple delays.
Approach
CVM continuously converts temporal margins on dependency edges into proportional velocity commands and propagates an exponentially damped slowdown signal along the dependency chain through self-recovery, direct cushioning, and cascade propagation steps.
Key results
- ~25% average makespan reduction in simulation across 5–8 agent scenarios
- ~35% makespan reduction in real-world trials with eight e-puck2 robots
- Maintains collision-free execution under simulated and real agent malfunctions
- Outperforms binary go/wait baselines across varying team densities
Why it matters
Provides a practical, low-complexity execution layer that significantly improves fleet throughput and robustness for real-world multi-robot deployments.
Abstract
Multi-Agent Path Finding (MAPF) plans are in- creasingly deployed on real multi-robot fleets, where commu- nication dropouts, actuator faults, and sensor noise routinely cause individual robots to deviate from the planned trajectory. We propose Cascading Velocity Modulation (CVM), a contin- uous execution controller that maps the temporal margin on each dependency edge into a proportional velocity command and propagates an exponentially attenuated damping signal along the dependency chain. CVM runs a three-step control loop: self-recovery, direct cushioning, and cascade propagation. CVM reduces the makespan by about 25 percent on average compared to a binary baseline, over ten randomized scenarios with 5 to 8 agents, each containing a malfunctioning agent that suffers an unexpected delay. An experiment with eight e-puck2 robots reproduces about a 35 percent reduction under two simultaneously malfunctioning agents.