Biologically inspired transportation: a distributed intelligent conveyor

点击翻译按钮获取中文摘要

A parallel manipulator is a massive array of simple individual actuators with a small power density that collectively transport and position objects with masses considerably higher than the force generated by a single actuator alone. This design is inspired by the biological phenomena of cilia, small hair-like structures on the surface of cells which can either sense local properties such as in the rod photoreceptors for vision or in olfactory neurons for smell, or can move in coordinated wave action to move liquid over their surface, as in the trachea and kidneys. Employing these capabilities in an analogous array of micro-actuators will produce a conveyor of parallel intelligent manipulation able to sense object properties, move them in different directions and effectively sort objects according to their properties. Crucially, the actuator array will be capable of communicating local information about objects to other parts of the array to enable coordinated action.Parallel intelligent manipulation plays an increasingly important role in intelligent robotics, computer science and intelligent manufacturing systems. Significant advantages of the distributed manipulating system are task flexibility (it can be dynamically reprogrammed to implement another task); massive-parallelism (it can process several different objects simultaneously, and different parts of the manipulator can perform separate tasks concurrently); fault tolerance (faults in single actuators do not restrict performance of the system as a whole, so operation of the system can be maintained); autonomy; and the ability to process objects simultaneously and independently.The overarching aim of the project is to build an intelligent autonomous massively parallel manipulator for distributed sensing, recognition, analysis, sorting, transportation and manipulation of light-weight objects. A paradigm of reaction-diffusion computing, i.e. information processing and computation by spreading wave-patterns in non-linear media, will be employed in the control system of the manipulator.Using evolutionary computation and machine learning, we will develop new principles and implementations for non-linear medium based control, and introduce a range of algorithms for distributed sensing (of object properties such as shape), filtration (sorting different objects according to common characteristics), orienting (ensure objects are facing and moving in the correct direction), positioning (moving objects into the correct path of travel on a different part of the manipulator) and shape-determined transportation of the objects.This manipulator system not only has the potential to impact upon the academic community in terms of the advancement of evolutionary algorithms, reaction-diffusion computing, and intelligent robotic systems, but also has ready application domains in industry such as high-tech manufacturing, enabling an advanced network of sensors to control dynamics of mechanical components, automation of assembly of nano-devices, and medical applications such as prostheses and computer controlled implants.