Structural health monitoring plays a major role in maintaining large critical infrastructures like bridges, breakwaters, dams, gas and water supply networks, and transport pipelines, which in general require complex and expensive routine inspections and maintenance procedures. Most of these structures are exposed to harsh environments and heavy loads and some of them (like rubble-mound breakwaters) are designed, due to their characteristics, under the proviso that maintenance and protection works will certainly be required during the structures life. The cost of the structure, its expected behavior, as well as the consequences of its failure, do completely justify the existence of a monitoring program, which will help in the decision making process relative to optimal timing and extension of maintenance, or even repair, works. This process should be based on the structure diagnosis, which, in turn, should rely on a set of state variables that clearly characterize the health of the structure.

Accurate health monitoring and diagnosis of critical infrastructure will increase the efficiency of maintenance and repair plans, with inherent benefits in terms of cost reduction and damage minimization in case of disaster. This presents an opportunity for the development of advanced robotic surveying tools, namely uninhabited aerial vehicles (UAVs) equipped with state of the art laser, multi-spectral and hyper-spectral remote sensing devices, high accuracy inertial platforms, and positioning systems. These vehicles should be able to perform high accuracy tri-dimensional surveys of structures with the objective of producing, in real time, accurate data sets with the required spatial and temporal resolutions and thereby providing quantitative information vital for a well-founded diagnosis.

Recent advances in sensor technology and the increasing availability of computational capacity are steadily affording UAVs higher degrees of robustness and reliability in challenging operation scenarios, taking place in uncertain and possibly remote environments. Unlike fixed-wing aircraft, helicopters were designed to execute vertical flight maneuvers, including hovering and vertical take-off and landing (VTOL). The trade-off for such maneuverability is an inherent complexity that translates into a highly nonlinear and unstable dynamical system with wide parameter variations over the vehicles flight envelope.

Motivated by the foregoing considerations, the aim of this Project is to develop an Autonomous Helicopter specially tailored for critical infrastructure monitoring by means of collision avoidance mechanisms and absolute and sensor-based navigation and tracking control laws, which rely on the aircrafts advanced sensing devices and exploit the properties of the configuration space to express the dynamics of flying robots, that is the special Euclidean group SE(3). In preparation for future monitoring and inspection scenarios that can require the use of multiple helicopters equipped with complementary sensing devices, additional research effort will be placed on the area of cooperative control of multiple Autonomous Helicopters.

The development of such a system involves a wide range of research topics, including dynamic modeling and identification, navigation, guidance and control, real time systems, and mission control. This project team has already addressed some theoretical aspects of these topics within the scope of previous projects. Based on the work formerly developed, the current project will focus both on developing the experimental components and on extending the theoretical results previously obtained within the fields of guidance, navigation, and control. The resulting UAV will be equipped with a distributed real time computing network, a reliable wireless communication system, and sensing devices. Given the envisaged applications, the latter include inertial sensors, a GPS, a laser range finder, and a camera array composed by a digital video camera and an infrared camera.

The camera array will be mounted on a pan-tilt unit. To reject the low frequency oscillations induced by the vehicle and stabilize the cameras image, a closed loop control system will be implemented whereby the pan and tilt motions of the camera are compensated based on image data and inertial information provided by the aircrafts navigation system. It is then possible to direct the camera array to a specified target and ensure that it keeps a steady image, regardless of the pose assumed by the helicopter while maneuvering or even under wind induced disturbances.

The research work will focus on bridging the gap between theory and practice, by taking into account actual characteristics of the systems at hand in the development process. Evaluation of system components reliability and of overall performance will be carried out in a controlled environment resorting to Hardware In-the-Loop Simulation to reduce both the number of required field trials and their associated risk factors.