Mohsen Asadnia - Artificial Intelligence and Data Science in Environmental Sensing

In recent years, increases in industrial residue have become a significant environmental threat. These residues can cause problems for natural ecosystems and their inhabitants, including animals and humans. Environmental monitoring through sensing is one approach to predict or detect the presence of pollution from such residue. One of the emerging tools to do distributed sensing, and thus a novel approach to environmental sensing, is to use swarms of robots to carry sensors. Swarming robots are programmed to move like ants, birds, or other swarming or flocking organisms to distribute through and explore an environment. Swarm robots have advantages over other multiagent or single-agent systems because of their number and decentralized control strategies, which means they can carry multiple sensors and explore wide areas in a fault tolerant manner. However, there are challenges that remain before swarm robots can be applied productively in this scenario. This chapter addresses two of these challenges: (1) how swarming behavior can be achieved quickly on a given set of robots and (2) how the swarm can conduct environmental sensing. We present a novel system design that combines two algorithms: the first is a novel algorithm for autonomous tuning of swarm behavior and the second for conducting an environmental sensing task, represented as an area coverage problem. We show that the proposed system can tune the behavior of swarms, suitable for completing a coverage task more effectively than an untuned group of robots. We demonstrate the system in both point mass simulator and in simulated robots.

1. Their autonomous movement reduces the need for human involvement in this task.
2. The spatial and temporal changes of the environment can be better addressed by a swarm of robots than what can be done by human or other static sensors. This can be achieved by AI algorithms to optimize the environmental sensing, number of robots, and their distributed sensing.
3. The sensing task can be done more accurately and with less energy consumption by the aid of coverage, mapping, and path planning solutions.
4. Swarm robots can solve environmental sensing quickly, in cases where fast detection of contaminant resources is crucial.

• 1. Collision avoidance (repulsion): following this rule, each boid avoids collision with the other boids.
• 2. Velocity matching (alignment): based on this rule, each boid should attempt to match its velocity with the other boids moving in its neighborhood.
• 3. Flock centering (attraction): regarding this rule, the boids should attempt to move close to the other boids within their neighborhood.