ASP
The rising share of e-mobility in the transport sector motivates this project. Electric vehicles, whether battery or hydrogen driven, are increasingly replacing combustion-based vehicles. Nevertheless, the absolute number of electric vehicles on the road is still relatively small. This is partly due to the fact that the infrastructure for re-charging is only slowly increasing, and partly due to the relatively high start-up costs of such vehicles. Making electric vehicles affordable requires the production costs to be lowered, which needs among other things to reduce the production cost of one of the main cost drivers: the electric powertrain. The electric powertrain consists of e-motor, gearbox and power electronics. Producing this electric powertrain at as low costs as possible is therefore a key requirement for future automotive development.
Reducing production costs requires changes in the production paradigms, in particular when compared to production of conventional powertrains. It is not sufficient to tailor existing manufacturing processes for increased efficiency. This is due to the fact that the current production lot size of electric powertrains is still at low volume meaning that the full capacity of the production machines cannot be used. Producing different electric powertrains in the same production plant solves the capacity problem, but new challenges emerge: How can different electric powertrains be produced in the same production plant without any delays, ramp-up time and defect parts? The answer is a new paradigm: adaptive and cognitive production.
Goals
The goal of this project is to establish a new paradigm of production systems for electric powertrain assembly. Future assembly lines for electric powertrains must be (i) more flexible, to achieve assembly of high variety and low volume parts. Combining the high variety with high efficiency addresses the issue of ramp-up time (converting the assembly process). (ii) Reducing ramp-up time is essential when assembling many different types of powertrains. In this project, reducing the ramp-up time will be investigated by combining simulation of the assembly process (virtual) with data from the real assembly process (physical). This combination will lead to a very significant and powerful prediction and better plannable ramp-up time. Decreasing production costs can also be achieved by (iii) reducing assembly time. This can be realised by (iv) well-balanced human-machine interactions in each assembly cell and assembly operation. Regarding this, the human factor (cognitive load in complex assembly process) needs to be considered. Hence, the cell itself needs cognitive and self-learning elements for engaging flexibly with the human worker. Flexibility is also a factor when connecting different assembly cells. Flexibility in terms of connected assembly cells will be investigated for (v) adaptive and flexible plant and cell layout structure.
Approach
Starting with the analysis of the architecture of different electric powertrains and their assembly processes, we will pinpoint similarities and differences. Based on this analysis, the requirements of the whole assembly line and each constituent assembly cell will be established. Existing technologies for each requirement (e.g. collaborative robots for heavy parts, cognitive guidance systems for complex assembly operations) will be investigated in more detail. After evaluation, a candidate solution will be implemented in an existing physical assembly process and in a simulation-based model. This proof of concept will present new adaptive and cognitive paradigms in production, as well as an overview of the weaknesses and challenges of this new paradigm.
Expected and Achieved Results
Starting with the analysis of the architecture of different electric powertrains and their assembly processes, we will pinpoint similarities and differences. Based on this analysis, the requirements of the whole assembly line and each constituent assembly cell will be established. Existing technologies for each requirement (e.g. collaborative robots for heavy parts, cognitive guidance systems for complex assembly operations) will be investigated in more detail. After evaluation, a candidate solution will be implemented in an existing physical assembly process and in a simulation-based model. This proof of concept will present new adaptive and cognitive paradigms in production, as well as an overview of the weaknesses and challenges of this new paradigm.
