In order to visually inspect large structures for damage, typically working scaffolds are installed or specifically trained workers need to climb on the machine or structure at periodical intervals: taking a lot of time and money for material and personnel. Moreover, machines must be down for the inspection, resulting in indirect downtime costs. Visual inspection by workers has the disadvantage that potential damages are experienced subjectively by the person that is performing the task. To classify the degree of severity of a defect, the evolution of size and shape of the damage is highly relevant.
Crane systems in steel plants, for example, are exposed to heavy soiling and high wear. Smooth production processes require regular inspection of all components such as runways, ropes and hoists. These areas are often difficult to access due to their location at great heights. At Salzgitter Flachstahl GmbH in Germany, for example, there are about 40 kilometers of crane runways at a height of 12 meters with 350 cranes distributed on the plant. Around 80 percent of these runways are located indoors. ROBOINSPECT will develop a robotic system that solves those problems by using an unmanned aerial vehicle (UAV) which is specialized for the inspection of plant facilities at great heights in combination with maintenance software based on intelligent image evaluation algorithms. Concepts for autonomous navigation will even allow the application of the drone during operation and makes downtime due to inspection obsolete to some extent