Robotic vision has been forecast for decades, but in reality machine vision does not yet work very well. I suggest using a new approach to image analysis, similar to what evolution has done for mammals. If machines and robots could see the way mammals do, then it would be much safer to work with them, have them perform work that is too dangerous for humans, and in general augment human vision with machine vision. Elder care robots, automotive guidance systems, drug discovery, fMRI image analysis, even air traffic control would benefit very substantially from machine vision that is real-time capable and fully deterministic. One such system is being developed at Parimics, and the simulation models show that our system works as well as we had anticipated. We have made great progress in the implementation of the chipset, and we now look into the embedded software and host software to interact with. With the proposed chipset we'd make the world a safer place for humans. I would like to use the funds to accelerate chipset development and more importantly, build an FPGA-based model that can be used to show its effectiveness so that interested parties can start their own designs based on that chipset. Ideally, an ecosystem would evolve around this, creating a new industry of real-time and deterministic machine vision systems.