FleetScape: Mixed reality sandtable lets one pilot manage 15 drones
A new MR system reimagines swarm supervision as spatial interaction, not scaling up single-drone controls.
As autonomous drone fleets grow from single units to coordinated swarms, the operator's role shifts from pilot to supervisor—but existing interfaces treat multi-drone control as just a scaled-up version of single-drone operation. FleetScape, developed by researchers at CNRS, Université Paris-Saclay, and National University of Singapore, challenges that assumption. It is a mixed reality (MR) sandtable that externalizes real-time mission, safety, and environmental data across layered spatial representations. Operators can fluidly toggle between manual intervention and autonomous supervision.
The team built a high-fidelity building inspection simulation that streams synchronized multi-drone and environmental data into the MR visualizations. In a user study with six experienced drone pilots managing fleets of up to 15 drones, FleetScape supported situational awareness through its spatial layers and clarified when switch between control modes was happening. However, researchers observed a limit to situational awareness as fleet size increased, leading pilots to adopt different supervisory strategies. The study concludes with design implications for building scalable drone fleet supervision systems.
- FleetScape uses a mixed reality sandtable to visualize layers of real-time mission, safety, and environmental data for drone swarms.
- In a simulation with 6 expert pilots managing up to 15 drones, FleetScape improved situational awareness and control mode clarity.
- Awareness degraded with larger fleets, prompting different supervisory strategies and guiding future design for scalable swarm control.
Why It Matters
FleetScape points toward practical MR interfaces for human supervision of large drone swarms in real-world inspections.