Machine learning and computer vision have the potential to significantly improve the automation and autonomy of many industrial applications (e.g. offshore, automotive, telecommunication, gaming and multimedia) by enhancing the operational performance, decreasing cost related to manual operations, increasing benefits, minimizing losses, optimizing productivity and improving safety and security. The best-intuitive industrial application for an autonomous and automated system is process monitoring, where camera systems can permanently inspect for integrity of process operations. Examples include offshore drilling process and high-yield in metal production without human intervention for safe and high-performance and low downtime industrial operations. However, many natural industrial environments still present major challenges for full automation and autonomy because of changing environment conditions, e.g. challenging illuminations, difficult weather conditions and the highly dynamic environment, reducing visibility and safe operation without human intervention. While recent scientific advances in sensing, learning and computing aim to reduce the gap between scientific findings and their practical industrial deployment, the transfer of scientific approaches into practical and robust industrial solutions is not straightforward.
Project leader: Ahmed Nabil Belbachir
Category: Teknisk-industrielle institutter
Institution: NORCE NORWEGIAN RESEARCH CENTRE AS