The most costly quality issue in electric vehicle manufacturing often goes undetected during the defect screening stage. Such parts arrive on time, in sealed boxes, and with proper documentation.

In Tesla’s powertrain procurement department, production delays frequently occurred due to incorrectly labeled parts from suppliers. Two components differing only by revision level or material would often arrive with swapped labels. The systems would scan the barcode and accept it as accurate, so the error was only discovered on the production line.

Typically, these situations were addressed by introducing extra incoming inspections: increasing the number of boxes checked or adding more staff to the receiving process. However, this only treated the symptoms, not the root cause.

When a mislabeled part reaches the warehouse, payment for delivery has already been made, the entire order is with the recipient, and the production schedule is built around these parts. Inspection costs do not decrease—they grow in proportion to the volume.

Principles of Quality Management

According to the third principle of quality management, quality should be built in at the source, not ensured by inspection. The problem was that the source was at remote factories not directly managed by the company.

To solve this, an error-proofing (poka-yoke) system was implemented directly into the labeling and packaging processes at four suppliers. These were simple solutions, without machine vision or extra inspectors. For example, the label printer would not start printing until a gauge confirmed the physical feature distinguishing the revision. In another case, the part had to be weighed before the label could be printed. Colored stripes corresponding to the revision level were applied to parts so that mixed pallets could be spotted from a distance. The equipment cost was just a few hundred dollars per station.

After these measures were introduced, downtime due to mislabeling stopped. Similar methods were used to eliminate errors in quantity and kitting.

Criteria for Error Prevention

Not every supplier error requires new equipment to solve. Three questions help assess the situation:

• Is the error binary (yes/no)?
• Does it remain undetected in subsequent steps?
• Is the cost asymmetric (cheap to prevent, expensive to fix)?

Labeling and identification errors meet all these criteria. For the supplier, it’s just a mislabeled box; for the manufacturer, it means a halted line and extra costs.

Supplier Collaboration Recommendations

Suppliers were offered solutions that benefited them as well: reducing defects, rework, returns, penalties, and urgent shipments. Controls that save the supplier money tend to remain in place even if management changes. One supplier agreed only after being shown data on line stoppages caused by their parts.

Implementation of controls took place at the supplier’s site with their operators involved. A company engineer installed the equipment, and the operators contributed improvements to the design and workflow. Such devices are rarely disabled if the staff helped create them.

Control procedures were documented in materials the supplier already provided, such as the PPAP control plan, to ensure continuity through staff changes and annual revalidations.

Actual effectiveness was monitored via label scan logs, first-pass yield rates, and monthly photos of the station, which provided more insight than quarterly audits.

The Role of Inspection and Error Prevention

Machine vision and extra inspectors are sometimes necessary, but they should not be the first line of defense. Inspection is a recurring cost, and its effectiveness depends on human factors. Devices that make errors impossible are a prevention cost: you pay once, and the benefit lasts.

If the failure is binary (correct or incorrect part, matching or not matching label, right or wrong revision), it should be prevented at the source using key geometry, printer interlocks, or weight checks. Machine vision systems are better suited for tasks requiring measurement—such as dimensions, surface quality, or welds. Human factors are less reliable than automated devices.

Practical Conclusions

Error prevention is a crucial discipline in manufacturing. The value stream map should include the next tier of suppliers, since their labeling processes become part of your production process.

When a supplier error causes a line stoppage, it’s not always best to tighten warehouse controls. It’s more effective—and often cheaper—to address the root cause at the source, rather than holding follow-up meetings about the same issue. When an error becomes impossible, it disappears from the process.