Conference Agenda

In the context of product development within the automotive industry, physical prototype vehicles with new functions and components need to be built for testing and validation. The assembly of prototype vehicles is carried out through a manual process without automated assembly activities, thus requiring highly skilled workers capable of manually assembling vehicles. It is conducted stationary on lifting platforms, where the vehicles are mounted during assembly. A master craftsperson is responsible for coordinating the assembly of the assigned orders. Therefore, they allocate the vehicles to lifting platforms, schedule their assembly activities, and allocate the workers to them. Since prototype parts often undergo last-minute changes, not all required components are usually available at the beginning of the assembly process. Therefore, the assembly may be started even when not all parts are available to avoid delays in subsequent testing. Typically, the vehicles remain on the lifting platforms from the start of their assembly until completion. However, the master craftsperson can optionally decide to remove a partially assembled vehicle from the platform and allocate this resource to assemble another vehicle.

We model this problem as a resource-constrained project scheduling problem with multiple (vehicle) projects (RCMPSP). We use optional activities to model the mounting and unmounting of vehicles onto lifting platforms, effectively splitting the projects. Notably, our approach is the first to integrate decisions on project splitting into RCMPSP approaches. Based on numerical experiments, the potential of project splitting is assessed.

Automotive manufacturers have improved supply chain efficiency by introducing lean part delivery methods like just-in-sequence. Approaches such as the pearl chain concept (or in-line vehicle sequencing) take this even further by outsourcing part sequencing to suppliers. To facilitate this, the sequence on the final assembly line is determined several days in advance. However, the necessity to adhere to already communicated order sequences poses challenges for production planners in case of unexpected supply shortages. Sequence changes are restricted by warehouse storage and handling capacities because all parts delivered in sequence must be rearranged. An interim solution is to assemble vehicles without the missing parts and install them later in a rework step. However, this mechanism is limited by the manufacturing capacity for rework and the space available for storing incomplete vehicles. This frequently implies that resequencing or rework are no longer feasible resulting in costly production shutdowns. We introduce a car resequencing approach in pearl chain-controlled factories. The core idea is to cut segments from the pearl chain based on their exposure to supply shortages while considering the delivery lot sizes in which parts are supplied. This rescheduling of segments instead of single vehicle orders helps reduce the effort for part rearrangement. Simultaneously, we maintain a high production capacity utilization through the controlled deployment of rework and by pulling vehicles without shortages forward. We show the viability of our concept on a dataset of real vehicle sequences, missing parts, and lot sizes for part supply.