Machine vision has been a common tool in the fields of robotics and automatic machine control in order to allow repeatable processes to become a dynamic component of production processes. Omnirobotic has used depth vision developments with novel sensor fusion techniques to create a 3D Perception technology that instead allows machines to see and interpret part shapes and positions in real process time using a Digital Twin environment. By allowing to machine to interpret different types of parts and their positions, and subsequently adding process know-how for tasks like painting or sandblasting, a robot can finally conduct processes autonomously in high-mix environments without the need for manual programming or precision fixturing - all because this package of technologies eliminates their incremental costs as part variation grows.
This talk will discuss the development of manufacturing and 3D Perception, explaining the history of machines and sensors used to automate industrial processes. It will explore how sensor fusion in the industrial space functions much in the same way as it does for self-driving cars - by generating a 3D proxy of the real world environment in which a robot operates. 3D perception on industrial parts must also rely on the decomposition of parts into geometrical primitives like face, edges and summits. Real 3D perception must also figure out how those primitives are hierarchically organized, a concept often referred to as semantic geometrical representation.
Ultimately, this permits a new approach - autonomous manufacturing - which allows manufacturers to focus on design and delivery of products while further automating their actual production through value-added processes. These systems on their own can deliver safety, quality and productivity improvements while eliminating waste and rework. Autonomous manufacturing allows use of robots in High Mix manufacturing environments that, in turn, also permit manufacturers to increase their overall capacity and add flexibility to their processes with equipment that is commonly available today.
