Researchers at the Harvard John A. Paulson School of Engineering and Applied Sciences have discovered that adding a controlled amount of randomness to robot movements can prevent traffic jams in crowded environments.
According to ScienceDaily, the study demonstrates that while adding more robots to a task can initially speed up progress, too many units in a limited area eventually cause a standstill.
The research, led by applied mathematics Ph.D. student Lucy Liu, suggests that allowing robots to 'wiggle' slightly instead of following rigid straight lines helps them slip past one another.
Finding the Goldilocks Zone
Using computer simulations, the team modeled 'agents' moving toward random destinations. They found that perfectly straight paths led to dense clusters and total gridlock, while excessive randomness caused inefficient wandering.
However, a 'sweet spot' exists where a specific level of movement variation—referred to as 'noise'—allows the system to maintain a steady flow.
"This might be counterintuitive, because how could randomness make things easier to work with?" Liu said, according to the report. "But in this case, when you have a lot of randomness, it becomes possible to take averages -- average distances, average times, average behaviors. This makes it a lot easier to make predictions."
The team developed mathematical formulas to estimate the 'goal attainment rate' and determine the ideal balance of density and randomness to maximize performance.
To validate the simulations, the researchers conducted physical experiments using small wheeled robots in a lab setting. Working with physicist Federico Toschi at Eindhoven University of Technology, the team used overhead cameras to track robots carrying QR codes.
Despite the physical robots moving less precisely than the digital models, the experimental results mirrored the simulated patterns. The study suggests that effective group behavior does not require centralized control or complex intelligence, but can emerge from simple, local rules.
