700470Perceptual RobotsClosedGRD Proof of MarketInnovate UKThe project intends to assess the viability of a revolutionary methodology for developing autonomous systems with significant impact potential on the world of robotics. Currently, the primary robot building approach is based upon modelling observed human behaviour. Robots simply mimic human movements, with no internal purpose. Various examples exist, from robots modelling dancing movements to vacuum cleaners operating with minimal human input. While this approach works in highly constrained areas (e.g. car assembly lines), there is a growing realisation of its limitations, and invalidity, due to the complexity of solutions needed to meet users` needs. Complex maths is involved and robots can only perform tasks they were designed for, with limited ability to overcome uncertainties in both the model and environment. Perceptual Robots (PR) represents a much simpler cost-effective approach overcoming the constraints presented by conventional wisdom. Inspired by Perceptual Control Theory, the proposed approach translates humans’ purposeful, goal-oriented behaviour into the robotics field. Rather than modelling human behaviour, the framework models perception-based processes that enable experts to develop artificial systems with internal purposes; able to conquer dynamic and unpredictable environments by controlling high-level, abstract perceptions. Neither complex maths nor physics knowledge needs to be modelled. The system`s behaviour, arising from a hierarchy of perceptual controllers rather than from the stimulus-response model, is a real step-change in the industry, allowing experts to use more sophisticated robots capable of performing more complex tasks with significantly less processing power. The project will identify technology issues occurring within the industries exploiting robots and will assess the potential PR has to resolve these problems. It will help to verify which markets are most viable and uncover the maximum potential for this idea