The advance of the Internet of Things into the industrial environment is just beginning. A lot of research and development still needs to be done. Nevertheless, a number of projects are already demonstrating the prospects for adding value that Industry 4.0 opens up.
At the last Hannover Messe, the world’s largest industrial trade fair, the German Research Center for Artificial Intelligence DFKI demonstrated how manufacturing in the age of the Internet of Things might look: Its demo plant presented a complete production line for keyrings fitted with miniature torches. The production process begins in the picking station, where the keyring is assigned its individualised production data and production-specific parameters. The product carries this information on an RFID chip throughout its life-cycle. During the production process, information is exchanged between the controller and other automation components such as robots, the machine tool control, etc. A robot transports the workpiece from station to station along the line. The production status is updated in the intelligent product memory between the individual work stages. The “smart” communicating products and machines on the line are autonomous elements with a local control intelligence – so-called cyber-physical systems (CPS). They also enable custom product variants to be manufactured flexibly and efficiently in small volumes.
Networking prevents failures
Experts believe it will take another 15 to 20 years for such self-organising dynamic factories to become part and parcel of the manufacturing landscape however. But if a “smart factory” can be defined as a self-controlling production system capable of responding autonomously to changes or disturbances and initiating the appropriate actions, then such elements of the Internet of Things are indeed already to be found in everyday industrial applications. One example is predictive maintenance, whereby machines undergo maintenance before a malfunction actually occurs. This utilises the networked interaction of machinery, products and other systems involved in the production process. Sensors are also employed, to monitor states such as the temperature or vibration of machine components. That data is combined with data from third-party systems, such as enterprise software, and analysed. In this way, patterns indicating potential malfunctions can be identified in good time and appropriate action taken. An example of this is the system and services platform from Bosch Software Innovations. It integrates all systems and objects as well as all functionality from data analysis through to rule-based process activation.
Intelligent components already exist
The kind of “intelligent” sensor–equipped components needed for predictive maintenance do already exist. Swedish roller bearings, seals and lubricant systems specialist SKF has developed a bearing capable of continuously signalling its operating conditions. It incorporates a sensor package which continuously records data such as rotation speed, temperature, velocity, vibration and other parameters. The bearings utilise the application environment to generate the energy they need to perform their functions. They are also equipped with a wireless communication module. The bearings communicate wirelessly among one another via an interface. In doing so, they act as nodes of a tight-meshed network, passing relevant state data from one bearing to the next until it reaches its destination. So instead of an “Internet of Things”, an “Internet of Bearings”…