Abstract
An innovative and flexible approach is introduced to address the challenge of self-organizing a group of mobile robots into cubic-spline-based patterns without any requirement of control points. Besides the self-organization of mobile robots, the approach incorporates a potential field-based control for obstacle/collision avoidance. This will offer more flexibility to swarm robots to efficiently deal with many practical situations, including smoothly avoiding obstacles during movement or exploring and covering areas with complex curved patterns. Essentially, this challenge is approached by proposing a formation control model based on a smoothed particle hydrodynamic estimation technique, which uses special cubic-spline kernel functions applied here to interpolate the density of each robot in the swarm. The moving information is used to weigh the distances to the robot's neighbors available in its field of view. Then, an artificial physics mesh is finally built among each robot and its three available neighbors having the smallest weighted distances. Significant results toward emerging cubic-spline patterns are shown with a swarm of foot-bot mobile robots simulated in the ARGoS platform. Analysis results with different metrics are also conducted to assess the performance of the model with different swarm sizes and in the presence of sensory noise as well in the presence of partially faulty robots.