An eye for accuracy: Visually validating assembly accuracy saves time and money for automaker

Taking the guesswork out of assembly accuracy can have benefits well beyond cutting costs. When Ford Motor Company implemented a 3D vision system at its engine plant in Windsor, Ont., the automaker increased assembly accuracy in nut-running, reduced rework time, reduced downtime and increased efficiencies. But arguably the biggest benefit the auto giant has seen since implementing the vision system is zero incidences of engine failure associated with loose bolts.

Ford manufacturers 3,600 engines at its 1.2-million-sq.-ft. engine plant each day. For a company that measures success by quality, just one faulty engine could result in the loss of business.

“Quality means so much in this competitive industry,” says Sara Stojcevski, a fastener engineer with Ford.
Previously, the company relied on PLC alerts to detect misapplied bolts on the engine mounting line. When a problem was detected, engines were diverted to a repair bay where technicians would manually unfasten and re-fasten all 16 bolts on the V-8 engine, even if only one bolt was loose. The system had to acknowledge 16 hits before releasing the engine.

But this process was unreliable, ineffective and time consuming.

“The problem with that is you can hit the same bolt 16 times and get an okay torque and [release the engine], but there could be 15 other bolts that are loose,” says Stojcevski. “That resulted in our customers receiving engines with loose connecting rod boltsÖSometimes they caught it at our vehicle operations plant, but we’ve also had many get right to our customer.”

The plant needed a solution that would pinpoint which bolt needed to be fastened and confirm that it was, in fact, fastened, eliminating any unnecessary unfastening and refastening of bolts that were already tight. Stojcevski contacted Ross Rawlings, president of Oldcastle, Ont.-based Radix Controls, and Nikola Dimitrov, the companyís vice-president of engineering, to come up with a solution.

The team tested various types of systems, including an ultrasound system and an ultrasound system in combination with an inertial navigation system, but neither was reliable enough for Ford’s application because of the excessive noise associated with manufacturing environments. Finally, it was decided that a vision system was the appropriate solution because the technician needed to see that the proper bolt was being torqued down.

Enter Tool Tracker, a 3D vision system developed by Radix Controls. The system had been in development for two years before being implemented at Ford.

How it works
The Tool Tracker is an integrated system for visually tracking and identifying the location and orientation of an object within a given space. It uses a target, one or more cameras and subsequent analysis within the mathematical model in the system software. The system also includes industrial computers to analyse the images, and a graphical user interface that provides operator feedback that the operation was completed.
Using a strategically placed camera, Tool Tracker validates that the operator with the torque gun is doing the correction at the right place, and that it has been executed as required.

The tool uses high-speed visual tracking technology to track a target permanently mounted on the hand tool (in Fordís case, a torque gun). It tracks the target and reports the toolís socket position. That information can be used to determine if the correct nut has been fastened. The system links with process data, validates the position and corresponding torque event for quality programs, determines the number and position of bolts that have been tightened, and verifies results to the operator, who then releases the engine.

Implementation and benefits
The installation took two weeks, and included mounting the Atlas Copco PowerMacs controller stand and the fastening tools bracket; making modifications to the existing engine rotate so the team could locate the engine; programming the PowerMacs controllers; electrical wiring and piping; calibrating the camera; teaching the bolt locations; interfacing the Tool Tracker HMI with the repair bay PLC (a Modicon 984); and error-proofing the repair bay PLC logic. Once everything was in place, the team needed a couple of days to configure the application.

The Tool Tracker is currently implemented in one repair bay at Ford’s engine plant, but Stojcevski says that there are other areas of opportunity that the company is looking into. She says the system could be implemented anywhere on the floor where error-proofing is required.

“We have shared this technology with our technological department within Ford, and they are extremely impressed,” she says. “They are thinking of buying the product and using it in other areas of Ford.”

It’s been almost a year since implementation, and the team is happy with the results. The system has reduced rework time by as much as 80 per cent and, because the technicians are doing fewer steps, it has allowed them to work faster, churning more engines out of the repair bay than ever before.

“We’d get instances where they were so busy in that bay, because our repair bay only holds so many engines. Once it fills up, the line will stop,” explains Stojcevski. “Because they’d have to do 16 hits before they can release anything, if there was any sort of quality issue with engines that were coming in, we would stop the line because [the technician] couldnít do it fast enough. So now with this, itís more efficient. Very rarely do we get backed up to the point of stopping the line. Actually, I donít even think we’ve had that issue.”

The Tool Tracker also automated the error-proofing aspect of the repair bay. Previously, the error-proofing technique was to have the lead supervisor verify that the engine had 16 hits. But the lead supervisor was responsible for the entire line, and was not always available, so they would have to wait. The Tool Tracker has automated that part of the job, and frees up the lead supervisor to do other things.

The system also provides traceability, since it collects historical data. If there is a recall, production records could be scrutinized rather than the vehicles, resulting in a reduction of the number of vehicles brought back for examination, which will save money.

There are additional impacts on the bottom line. Stojcevski says that the automaker quantifies its engine exchange due to loose bolt failures at about $5,000 an engine if they have to exchange the engine. She estimates that the company has saved $150,000 by implementing a system that ensures that engines with loose bolts do not leave the repair bay, and thus never have to be exchanged.

“But the biggest savings,” she says, “is our customer satisfaction. We can’t put dollars to it, but it’s a big deal to us.

“When this system is in use, and used properly, 100 per cent of engines have now been good engines. I havenít seen any loose rod bolts come back like we did in the past,” she says. “So as you can see, our quality depends on this system.”

TOOL TRACKER AT A GLANCE

What it does: The Tool Tracker links with process data; validates position and corresponding torque for quality control programs; determines the number and position of bolts that have been tightened; and verifies results to the operator.

Features: The Tool Tracker can track with one or more cameras; it can track multiple targets simultaneously; it has static repeatability of one millimetre – how close several measurements are if the target is not moved; dimensional accuracy of one millimetre – the quality of the measurement obtained; it supports multiple resolutions; and it uses standard protocols.

Applications: Industrial (tracking exact placement of large machinery); automotive (accuracy and verification of assembly operation); packaging (automating pickup and placement of shipping boxes); and aerospace (docking applications).