Can humans and robots team up to map soil?

Deploying robot swarms in the real world is expensive and time-consuming. This work replaces most hardware with humans following directions from a smartphone app connected to ground sensors; a central server runs the swarm algorithm and coordinates their actions. The team validated a soil-mapping algorithm that assigns scores to locations and biases search toward promising areas, showing superlinear efficiency gains—it works equally well when humans are the "robots." This approach dramatically lowers barriers for testing swarm algorithms outdoors.