HUman-MAchines CooperatioN for flexIble production SysteMs – HUMANISM

An objective of the factory of the future is to reach a high level of mass customization, low level of energy consumption through more agility of the global manufacturing system. Such a demand implies to design more flexible and more resilient manufacturing system. One way is to involve new technologies from digital and manufacturing technology fields. But at the same time, the Human must be kept at the center of the global process for decision making and control. In this context, HUMANISM proposes to use a Human centered design approach. The main idea is to balance Human and technology involvement, taking advantage of automation advances and Human capabilities such as experiences and adaptabilities. But since IMS are composed of self-organizing entities, one may face emergent behaviors, leading to a source of unexpected and abnormal risky situations; HUMANISM ambitions to develop a methodology to design cooperative assistance systems to support Human awareness and decision making for IMS, as well as assistance systems to control uncertainties.

The Human centred approach adopted in the project will rely, on one part, on an integrated approach, the Cognitive Work Analysis (CWA). It is considered to be one of the main methods taking human factors into account in the design of sociotechnical systems. This method will here support the design of assistance systems in IMS according to Human current and future needs redefined by new technologies. On the other part, it will rely on Human-Machine Cooperation principles (HMC) that provides precise models to identify and implement the best organization and task sharing between Humans and new technologies. To ensure the genericity of our theoretical developments, HUMANISM will be applied on three different IMS using different technologies and models with a focus on their ability to self-organize the production: a cobot system, a swarm of intelligent products, and a swarm of AGV. What is interesting to consider is that each of these three concerned Artificial Self-Organized Systems (ASO) presents different adaptable or adaptive level of automation. They differ each other in the kind of interaction they have with Human (close vs. distant), and in their behaviour (predictive, with communication abilities and accessible data). Each ASO is already available in their real or simulated version in the laboratories involved in HUMANISM.