Using a wireless visual sensor network to harmonically navigate multiple low-cost wheelchairs in an indoor environment

• Proposed a harmonic multiple low-cost wheelchair navigation algorithm for an indoor environment.
• Adopted a penalty for potential conflicts based on wheelchairs׳ priority, velocity and distance between nodes to predict its travel time.
• Carried out extensive simulation experiments of different topologies in an indoor environment by employing three harmonic strategies.
• The proposed method is verified by implementing a realistic prototype.

Harmonic navigation of multiple low-cost robotic wheelchairs in a topology of wireless sensor nodes that are deployed in a dynamic and crowded indoor environment is a Non-deterministic Polynomial-time hard (NP-hard) problem. To address this problem, we propose a distributed multi-wheelchair global harmonic navigation algorithm. The distinguishing features of the proposed navigation algorithm are global search and local conflict resolution abilities. In the proposed algorithm, a travel time prediction method adopts a penalty for potential conflicts based on wheelchairs׳ priority, velocity and distance between the nodes. Moreover, three harmonic rules are proposed for: (1) giving the highest priority to humans, (2) giving the highest priority to wheelchairs, (3) giving flexible priority to wheelchairs. Through extensive quantitative simulations, we explore the performance of wheelchairs in various floor plan topologies and different values for the system parameters, and demonstrate that the properties of crowded indoor environments have important influence on the performance of global navigation, such as service time. The third harmonic rule establishes the trade-off between the performance of humans and robotic wheelchairs. At the same time, physical prototype wheelchairs are implemented and they verify the proposed global harmonic navigation algorithm. Some suggestions for robotic wheelchair designers, building architects and building owners are provided based on the conclusion of the experimental results.