Issue link: https://iconnect007.uberflip.com/i/1275152
58 SMT007 MAGAZINE I AUGUST 2020 Carlone and his students developed a representation of spatial perception for robots that is modeled after the way humans perceive and navigate the world. The new model, 3D Dynamic Scene Graphs, enables a robot to quickly generate a 3D map of its surroundings that also includes objects and their semantic labels, as well as peo- ple, rooms, walls, and other struc- tures that the robot is likely seeing in its environment. The model allows the robot to extract relevant informa- tion from the 3D map and query the location of objects and rooms, or the movement of people in its path. (Source: MIT News) MIT engineers are envisioning robots more like home helpers, able to follow high-level, Alexa-type com- mands. To carry out such high-level tasks, researchers believe robots will have to be able to perceive their physi- cal environment as humans do. "To make any decision in the world, you need to have a mental model of the environment around you," says Luca Carlone, assistant professor of aeronautics and astronautics at MIT. "For robots, it's a hard problem, where it's about transforming pixel values they see through a camera into an under- standing of the world." of ingredients in the recipe for each product is always required. Unexpected material shortages cause havoc, especially mid-way through a production run. Before any production commitment, availabil- ity of materials to complete the order must be checked, though with this being a fluid con- cept, it is not easy unless cages are set up to surround kits of materials that have been cre- ated. ERP, the tool purchased to manage this situation that we trust to keep our factory mov- ing, cannot help us because we are in between stock checks, with many unknowns in terms of actual physical inventory quantities. Software technology in manufacturing has radically advanced recently, specifically related to IIoT data exchange and associated digital MES functionality. This is a clear step up from the capabilities of ERP or incumbent MES sys- tems, though modern IIoT-based MES solu- tions are designed to directly complement ERP, as well as many aspects of PLM. With produc- tion processes linked up as part of the IIoT- based MES platform, every individual compo- nent that exists in the factory is accounted for, almost all without the need for human inter- vention—whether individually identified or supplied in bulk. The IIoT-based MES platform supports decision-making for immediate pro- duction changes, safe in the knowledge that the exact materials are definitely available for use without the indulgence of bloated stock levels. As the physical stock levels are so closely aligned with the system, the stock check pro- cess becomes virtual, continuous, without nasty surprises. ERP performs its work more seamlessly, optimizing the incoming supply- chain in terms of cost, risk, and quality. With such technology, we typically see up to a 75% reduction of materials on the shop floor. Short- term schedule changes are handled seamlessly without additional concerns. Warehouse sizes have been known to reduce by a factor of four when all the unnecessary bloated stock is removed. The overall effect on the bottom line of the business of the "trust in time" Lean supply chain within the factory is seen by many as being the most significant cost saving that can be made. "Trust in time" lean materials is a very simple concept and far simpler to computerize than the push-kit system that it replaces. It is a real differentiator that any operation, regard- less of size or location, can utilize today. SMT007 Michael Ford is the senior director of emerging industry strategy for Aegis Software. To read past columns or contact Ford, click here. New Model Aims to Give Robots Human-Like Perception of Their Physical Environments