Monitoring Structures with Teams of Mobile Robots

Abstract

This paper addresses the problem of monitoring structures with potential emergent damage through adaptive sensing provided by teams of mobile robots. Advantages of mobile robot teams for structural health monitoring include: 1. Multiple views of a given structure, 2. Adaptive movements that focus attention in response to observed conditions,3. Heterogeneous sensing and movement, and 4. Federated health monitoring and prognosis assessment through networked sharing and processing of information. Towards this end three cases of the use of mobile robot teams will be presented: 1. Heterogeneous robot teams for home and small building maintenance – Identifying, diagnosing and mitigating damage to homes and small buildings is a vexing set of problems for the owners. As an aid small controlled bristlebots and quadruped robot dogs (QRDs) carry sensors throughout a small building, assess conditions, provide prognoses and networked links to repair options; 2. Culverts are primary components of stormwater and flood prevention infrastructure. Inspecting small culverts is difficult for humans and large culverts are accessible but dangerous due to issues of confined spaces. Low-cost mobile robots have emerged as a competitive inspection option for accessible culverts with straight or short runs that permit wireless telemetry. Longer culverts and those with bends, branches and drop inlets pose challenges to the telemetry. Teams of robots extend the range of inspection through multi-hop video and control telemetry; 3. Ground penetrating radar (GPR) is a method of sensing subsurface infrastructure conditions with high-frequency electromagnetic waves. Conventional GPRs operate in a suboptimal monostatic or bistatic mode, are tedious to operate and have limitations in sensing congested utility subsurface conditions. Coordinated multi- static ground penetrating radar operated with mobile robot teams alleviates some of these concerns and provide better subsurface assessments with automated methods that focus attention on subsurface features of interest. Results from laboratory and field tests of these robot teams, as well as organizing principles of control and automated information processing are presented.