Linked data at the service of robotics

Smart data for the smart factory

“Industry 4.0” is a collective term (coined in Germany) for the concept of Cyber-Physical Systems, the Internet of Things and the Internet of Services, which together constitute the structure of a “smart factory”.

In a “smart factory”, cyber-physical systems monitor production processes and, through the Internet of Things, cooperate with each other and with humans, helping to make intelligent and decentralised decisions.

Evolving towards the Industry 4.0 model is vital in order to respond promptly to the needs of new production methods and of their growing complexity.

As identified by both academia and industry, there are 6 paradigms:

• interoperability: the ability of cyber-physical systems (such as conveyor belts or assembly stations) and humans to communicate and be connected through the Internet of Things;
• virtualisation: linking data from sensors (from the monitoring of industrial processes) with models and simulations;
• decentralisation: the ability of cyber-physical systems to make decisions autonomously;
• real-time capability: the ability to collect and analyse data in real time, making it possible to reduce reaction times;
• modularity: the ability of smart factories to adapt to requirements by replacing or extending individual modules.

Moreover, one of the key goals of robotics is to build intelligent machines able to communicate, collaborate and operate in a more natural and safer way. Improving robots’ knowledge and intelligence is vital in order to implement the Industry 4.0 model successfully.

Figure: a KUKA robot handling flat glass.

In this context linked data are the most advanced solution for:

• representing knowledge in robotics applications;
• integrating the data of automated systems;
• describing movements and tasks;
• supporting the recognition of objects and places;
• supporting human-robot and robot-robot interaction;
• multidisciplinary implementations.

Successful implementations already exist: LinkedData.Center has designed a product to build data models based on linked data and on Semantic Web technologies, extensible without limit.