Smart Industry: improving performance

Growing use of digitally available data in industry

AI in industry already uses specific cases where all data is digitally available, often coming from one source. Think of a vision camera monitoring the position of parts to be grasped by a flexible assembly robot as used in manufacturing and agro/horticulture.

Other examples are high-tech industrial sensors, which generate huge amounts of data with a lot of noise and where AI algorithms can improve detection accuracy. For example, dust particle detection in the semiconductor industry. 

In more advanced applications, TNO researches better digital twin models with multiple data sources and mathematical/physical simulation models. In those cases other aspects are taken into account. Think of data collection, cleaning, visualisation, sharing data, data sovereignty or confidentiality, cyber security and use of standards.  

Dealing with data challenges in smart industry

There are several challenges today in smart manufacturing and services. One has to do with using and combining multiple data sources (sensors, remote systems, vison cameras etc). We often face multidisciplinary or complex challenges that require extensive system modelling and subsequently, in practice,  the continuous changes in equipment configurations and product mixes result in the need to determine proper AI parameters using a minimum of data.

And for human-machine interaction there is a need for explainable, understandable or transparent AI to be used by operators in a manufacturing environment. In a cobot situation the operator doesn’t want the robot to behave unpredictably or too fast for the operator to adapt. 

TNO offers

• TNO is active developing AI based solutions in achieving radical goals as zero-defect, zero-surprises, zero-programming, etc.
• TNO is active in multiple Dutch public-private partnership Smart Industry field labs.
• TNO is promoting standards such as the International Data Spaces (IDS), Open Platform Communication (OPC-UA) and the use of open hard/soft/tool-ware for AI.
• TNO helps to develop policies for open data ecosystems for industry.