Machines must know where to go at the construction site. They also need at some points to be able to take a hold of objects with precision and lift them up safely. Indoor localization challenges have been resolved already for years using a computer vision concept called Simultaneous Localization and Mapping (SLAM). SLAM solves the task by simultaneously forming a map of the surrounding, possibly unknown, environment and at the same time localizing the crane on the newly formed map. Traditionally, SLAM has been implemented using expensive equipment such as light detection and ranging sensors (LiDAR). In addition, indoor premises have inherent characteristics that introduce challenges for computer vision, such as non-lambertian and textureless surfaces, and low lighting.
One of the research goals of the Artificial Intelligence for Industrial Vision (AIV) project, carried out by the Spatiotemporal Data Analysis research group at the Department of Computer Science, University of Helsinki, is to develop deep learning based SLAM method for enabling cost-effective autonomous crane operations. Aalto University’s Industrial Internet Campus, in a research laboratory simulating a genuine factory environment, where data has been collected for testing the method. The initial results are very promising and the research continues in developing the SLAM method further. The study may in the future also benefit the development of self-driving cars, since the same questions related to positioning and observing the surroundings concern them as well.
Laura Ruotsalainen, Associate Professor
Department of Computer Science, University of Helsinki
+358 50 556 0761