RICE – AI-supported process optimization simplified

Quickly leverage optimization potential

Machine learning can play a decisive role in identifying and leveraging optimization potential in industrial manufacturing processes. However, many ML projects currently fail due to high costs. This is why advanced tools and solutions are needed that enable time-efficient and cost-effective implementation.

As part of its self-financed technology development program, Fraunhofer IOSB has launched the Rapid Instrumentation and Control Environment (RICE) project. The goal is rapid (over)instrumentation for the maturation of machines and systems. This means that in RICE, we are developing hardware and software tools to collect system data in a minimally invasive manner while also actively intervening in processes—for example, to automatically explore parameter spaces and collect data series that are needed for training a process-specific AI model.

RICE environment in Karlsruhe and Lemgo

RICE was prototypically set up in the research factories in Karlsruhe (www.karlsruher-forschungsfabrik.de) and Lemgo (https://smartfactory-owl.de). It can be connected to existing systems as a retrofit solution. RICE sensors collect the process data, which is then processed by the RICE software. Actuators (including robots) can intervene in the processes depending on the application. The demo scenario is an injection molding process, which is particularly well suited for the use of machine learning methods, as numerous parameters that influence the result can be monitored (including temperature curves, mold filling, and motion sequences).

Pilot phase in metal and plastics processing

RICE is currently in the pilot phase and is being tested by two industrial partners—a metal processing company and a plastics manufacturer. At the metal processing company, the RICE system will help automate previously manual inspection processes, such as inspecting turned and milled parts, and carry out end-of-line inspections of bus door drives. At the plastics processing company, the drying process will be automatically adjusted depending on the residual moisture in the materials supplied. Another goal is to use automated quality testing to either reduce throughput time or increase the proportion of regranulate in the end product without compromising quality. In both pilot tests, the recorded environmental and machine data are used for wear estimation and adaptive machine maintenance.

The results from the factory environments help the RICE project participants further develop their demonstrators and the software. Their current goal is to prove the functionality of the prototype and increase the maturity level of RICE. This will make it possible in the future to use the RICE tools in industrial projects quickly and with minimal initial effort.