TOFAŞ – Mix Optimization Application

TOFAŞ is one of Turkey's largest manufacturers, working toward high-volume, uninterrupted and error-free flow in automotive production. Ensuring that the vehicle flow between the paint shop and assembly stations occurs in the right sequence and at the right time is critically important. However, variability during production—such as vehicles leaving the paint shop late or unplanned downtimes—could disrupt this order. This led to bottlenecks on the assembly line and faulty production. To solve this problem, we began a special optimization software development process with TOFAŞ.

Our team designed a special decision-support algorithm to deliver vehicles from the paint shop to the assembly line in the most efficient sequence.

Key elements of the system we developed:

• The rail transport system between the paint shop and assembly line was analyzed.
• Control of the switch systems at transfer points was made software-based.
• With special algorithms developed in Python, vehicles were routed to the assembly line in the ideal order.
• An Angular-based user interface provided operators with real-time control and monitoring.

Thanks to this infrastructure, the line between the paint shop and assembly—one of the most critical transition points in the production process—now operates with minimal human intervention.

In automotive production, vehicles arriving on the line in a mixed order create major issues because the assembly line must perform vehicle-specific equipment and operations according to sequence.

In this project we:

• Enabled on-the-fly re-optimization of vehicle order.
• Created a decision mechanism within the software that manages uncertainties at the paint shop exit.
• Automatically calculated and presented the appropriate sequence for the assembly line even under different delay scenarios.

As a result, sequencing problems on TOFAŞ’s production lines were significantly reduced.

The mix optimization software we developed delivered major gains in a very short time:

• Sequence operations previously done manually were automated by the system.
• Performance rate rose from 55–65% to 95–100%.
• Due to this significant improvement in measurement metrics, some management processes and targets had to be revised.
• Operational efficiency increased, error rate decreased, and the production flow became more predictable.