With increasing product variation and decreasing lot size (cf. mass customization trend), the assembly operator plays more than ever an important role at manufacturing companies of complex products. The Operator 4.0 paradigm adds more and more tasks, skills and responsibilities (e.g. teaching robots, inline checking of quality, analyzing data, sharing knowledge, collaborating with robots, etc.). Current operator support systems need to be combined towards a digital assistant which allows reactive and proactive application scenarios in order to maximize operator acceptance and performance.
In the reactive scenario the operator takes the initiative and asks the support system for additional information, supportive actions or feedback on his past, current or future tasks. In the proactive scenario the operator is observed and based on his behavior he is interrogated and gets suggestions for improving his work methods.
The creation of digital assistants for assembly operators is hindered by a number of technological barriers:
- Lack of a validated architecture which combines different assembly operator support services into a digital assistant;
- Unclear which assembly knowledge base representation is most suited for a digital assistant to connect upon;
- Unknown which context-aware multi-modal user interface would be most suited and accepted by an assembly operator;
- Lack of a validated classification engine (for question-response mapping) that could support an operator in his/her day-to-day activities;
- Lack of insights on how and to what extent conversational AI could be beneficial for such kind of digital assistant.
The goal of the OperatorAssist project is creating a digital assistant that is able to fulfill the role of Single Point of Contact (SPOC) towards an operator combining relevant support services on an assembly workstation, such as digital work instructions, human-robot task allocation, assembly quality monitoring, teaching ‘assembly sequences’ by demonstration, scheduling learning moments, performance assessment, human factors evaluation, etc. To overcome each of the technological barriers an approach is proposed where all modules from the conceptual smart assistant architecture are first explored separately on the InfraFlex and Smart & Agile Assembly reactive and proactive research cases. In a number of following iterations pairwise integrations are further detailed and finally a full integration of the smart assistant will be validated on 3-4 small validation cases.
The project targets utilization at two industrial target groups. On one hand manufacturing companies with operator-centered final assembly activities that want to improve their operator performance. On the other hand technology providers of operator support systems that want to improve operator acceptance and performance by evolving to smarter and more interactive tools (as a separate tool and as part of a bigger picture).
OperatorAssist_SBO is a Strategic Basic Research (SBO) project. We are looking for companies to join the User Group and work with us on the valorisation of the project.
Interested? Complete the form below and we will contact you as soon as possible.