AR-assisted real-time Quality Control for an Engine Assembly Line. 

Repetitive Tasks in an assembly line depend on muscle memory – but what happens when tedium sets in? 

The process of sequential torqueing is common across manufacturing units.  As the engine-head (in this case) is placed on top of the body, it’s important the worker follows pre-set order while torqueing the bolts in place.  Any variation is potentially detrimental and costs time/money to the OEM. 

Using head-mounted AR for real-time QC, and setting up a feedback loop. 

We designed an AR application for the MS Hololens, that used a marker-based tracking system to verify if the bolts were torqued in the correct sequence.  Any deviation immediately raised a visual cue to the user, with call-outs and guidance on the correct steps to follow. 

Designing a fail-safe by integrating AR with the existing torqueing gun on the Assembly Line. 

To further fortify this process, using a Raspberry Pi interface, the MS Hololens was paired with the torqueing gun, triggering an error/shutdown signal within 400 ms of detecting a deviation from standard practice. 

Analysis tells us that using an AR HMD, helps reduce errors by up to 80% over an extended period. 

The impact of integrating AR into their workflow was two-fold.  With real-time recognition, QC was more efficient, avoiding additional hold-ups down the assembly line.  Secondly, training the worker immediately on the correct process, reinforced what they’ve studied, further reducing the probability of an error in the first place. 

If you’d like to revolutionise the way your assembly line operators work, then do reach out.