Cloud-based measurement and data acquisition for industrial and cyber-physical systems.
Investigators: Prof. X. Xu, Dr. Lechler/Prof. Verl
In a broader sense, the data may be directly measured by sensors or obtained from a controller during interaction between a robot and soft tissue. There are two main challenges/foci this project faces. First of all, selecting proper sensors (type and specification) and retrofitting them onto the legacy systems, require a systematic approach. Secondly, considering various types of data, a seamless and tether-free method needs to be developed to manage data acquisition procedure and transfer of data to a central server. For data communications in this research, industrial protocols, like OPC UA, will be considered. At the moment, large data transfers across the Web are still inefficient. Many deploy REST (Representative State Transfer) services but RESTful services face a serious data transfer bottleneck issue when it comes to transferring large volumes of data. We will seek to implement FOREST (Fast, Optimised REST), a technique deploying a workaround which ‘tricks’ TCP connections into utilizing the faster UDT (UDP based Data Transfer), a hybrid protocol with speed benefits of UDP (User Datagram Protocol) and the reliability of TCP. The project aims to develop an optimized, low-latency data transfer protocol and investigate hardware-level reduction of latency.