The world has seen several manufacturing revolutions over the last few decades, from Henry Ford’s inception of the mass production manufacturing line to the onset of robotic automation.  Manufacturing is a process that has evolved yet a few core tenants and behaviors have remained steadfast throughout.

One behavior has been central to the manufacturing evolution: A part is manufactured, and then it is checked for quality.  That behavior has evolved slightly as demand for manufactured goods has grown and manufacturing has become a more mature process: Now not even every part is checked, and sample quality checks are performed here and there.  Over the last decade or so, we have pushed the process further by bringing some of the metrology lab to the automated production line, thus automating the “make-the-part-then-check-it” concept, but not really reinventing it.  Every CMM manufacturer has a list of projects where they have introduced a CMM to a fully robotic manufacturing cell or line.  This has generated some industry debates on what makes a CMM “shop floor” for this ever growing CMM niche.  As we progress iteratively, we are able to automate the inspection, creating a nice increase in productivity.  Set-up expenses and production floor real estate, however, may still be prohibitive enough to keep the QC lab off of the production floor.  The “make / check” mentality is largely still intact.

Time travel

What if we needed to drill a hole, but knew the hole was in the wrong spot before we drilled it?  We’d need a time machine to find this out, of course, but if we knew the hole was off before we drilled it, we could correct the error and never drill the bad hole.  Theoretically, we could approach zero defects.

If we look at manufacturing differently, we could get to a new manufacturing paradigm without a time machine.  There may be another way around the “make / check” behavior of the past.  Let’s rephrase the previous scenario.

What if we measured a hole location before we drilled it?  If we knew the hole was mis-located beforehand, we could simply adjust the location and avoid a future mistake.  Further, if we used external metrology equipment to measure a hole before hand, would we need to measure if again afterwards?

Can we do it today?

Let’s break this down a bit further: What would we need to know to be able to measure a hole before we drilled it?  First, we would need to accurately locate the piece or substrate that the hole is going to be drilled into.  Second, we would need to know where the drill head was in relation to the substrate.  Finally, we would need a stable, repeatable way to introduce the part to the drill.

If we could do this, we could make what seems futuristic and impossible a modern-day reality – but is it even possible?

Today, industrial robots provide a very repeatable method for reproducing a manufacturing task, and Nikon Metrology’s Optical CMM (OCMM) offers the ability to track locations of multiple frames simultaneously.  In the drill analogy, we could assign one frame to the drill and another to the part.  We could use industrial robots to introduce a part to a drilling station and locate the part to the location where a hole needs to be drilled.  Nikon’s Optical CMM understands where the part is and where the drill is in relation to part.  Using the drill (location) itself as the measurement probe, Nikon’s OCMM can actually measure the hole location before the drill command is executed.  More importantly, if the Nikon OCMM detects the part is out of position, it can send error corrections to the robots(s) to correct the hole location prior to executing a drill command. 

The spirit of what we are doing is simple: We are measuring before we drill and we use metrology equipment to assist manufacturing, not check it.  Currently, Nikon Metrology packages this concept as Adaptive Robot Control (ARC). By doing this, we are able to stop errors and eliminate that ancient “make / check” process.

Adaptive Robot Control (ARC) in Airbus UK

An example of utilizing this technology was done by Airbus UK.  The airplane manufacturer estimated they that drill 50 million holes per year, fifty percent of which are manual operations.  They required a design book tolerance of 0.2mm.  Their new aircrafts had 14,000 holes per wing with no chance of manual touch-up with zero tolerance for scrap wings.

With the use of Nikon Metrology’s Adaptive Robot Control, Airbus successfully completed a fully automated drilling station for the Airbus A340 D-Nose.  In addition to the Nikon OCMM running ARC, the cell incorporated KUKA, Gemcor, and Delmia technologies.

The cell uses two cooperative KUKA robots, which present the Airbus work piece to the Gemcor drilling station in the location where a hole is to be drilled.  In the Airbus scenario, the Nikon OCMM -  acting as an external metrology “monitor” - observed where the drill was to be placed prior to the hole being drilled.  If the Nikon OCMM determined that the hole was out of place, it could send corrections to the robots prior to the drilling sequence.  It was only after the Nikon OCMM determined that the drilling location was correct that the drilling sequence commenced.  Therefore, Airbus measured before drilling, not afterwards.

In the end, the addition of metrology- assisted manufacturing to the Airbus A340 D-Nose cell reduced assembly time in half.  It replaced 19 skilled employees (who were re-deployed onto the Airbus A400M manufacturing process).  Finally, in three years of operation, the cell has not miss-drilled a single hole.  In that time there has never been an unexplained issue or concession due to ARC.

Obviously, metrology-assisted manufacturing isn’t quite a time machine, yet it isn’t science fiction either.  It simply represents our ability to reinvent how we manufacture to get truly amazing results.  This process offers the ability to truly revolutionize a continually evolving process - something even Henry Ford would be proud of.