As electronics turn into smaller, quicker and extra advanced, the manufacturing programs used to construct them are present process a metamorphosis of their very own. Panasonic Connect is likely one of the firms serving to to drive that change.
A business-to-business know-how firm throughout the Panasonic Group, Panasonic Join offers sensible manufacturing options that mix surface-mount know-how (SMT) tools, robotics, manufacturing execution software program, AI-powered analytics and linked manufacturing unit platforms.
By means of its “Gemba Course of Innovation” technique, the corporate goals to attach folks, machines and manufacturing knowledge to enhance productiveness, high quality and operational resilience on the manufacturing unit ground.
Its applied sciences are utilized by electronics producers producing every little thing from automotive electronics and industrial tools to medical units and client merchandise.
As automobiles, smartphones and industrial programs incorporate ever-greater numbers of sensors, processors and digital parts, producers face mounting stress to put smaller parts with larger accuracy whereas supporting more and more versatile, high-mix manufacturing.
To debate these challenges, Robotics & Automation Information spoke with Scott Zerkle, affiliate director of technical operations at Panasonic Connect North America.
Zerkle works intently with producers adopting superior SMT applied sciences and has intensive expertise in manufacturing operations, course of optimization and manufacturing automation.
Throughout the interview, Zerkle explains why AI’s biggest worth immediately lies not in changing manufacturing unit staff however in supporting them by means of predictive upkeep, defect detection and higher operational decision-making.
He additionally argues that the following main step for sensible manufacturing won’t merely be including extra AI instruments, however connecting knowledge throughout machines, supplies and manufacturing processes to create factories that constantly be taught and enhance from their very own manufacturing historical past.
The dialog additionally explores how producers are adapting to more and more advanced electronics, why automation is changing into important for high-mix manufacturing, and the way the convergence of AI, superior sensing and robotics may reshape electronics manufacturing over the approaching decade.
Interview with Scott Zerkle
Robotics & Automation Information: Trendy automobiles, industrial programs, and client units now include way more sensors, processors, and digital parts than they did a decade in the past. How is that this growing complexity altering the way in which producers design and function manufacturing traces?
Scott Zerkle: Throughout the board, merchandise from client units to industrial tools carry way more electronics than ever earlier than. Trendy automobiles alone now depend on 60 to 100 or extra sensors, with some fashions exceeding 200.
That interprets to extra parts, tighter pitch spacing, and tighter placement tolerances than manufacturing traces had been initially designed to deal with.
Because of this, producers are dealing with challenges on a number of fronts. Placement accuracy and repeatability matter extra, as deviations that had been as soon as acceptable can now lead to defects.
An incorrect feeder load or an unverified changeover could cause actual disruption throughout extra part variants. That’s why we’re seeing extra emphasis on verifying supplies and settings earlier than a job even begins.
R&AN: As digital parts proceed to shrink in dimension whereas growing in functionality, what are the most important challenges dealing with producers in SMT meeting and PCB manufacturing immediately?
SZ: Element miniaturization is outpacing what older traces had been constructed for. We’re putting components with spacing measured in tens of microns, the place even the slightest deviation can throw off the solder joint.
And as parts turn into smaller and extra densely packed, printing and placement must get extra exact too, since paste deposition and placement accuracy should scale alongside them.
That’s pushing real-time correction, adjusting placement and paste quantity as situations shift mid-run as an alternative of operating a hard and fast program begin to end.
Printing, placement, and inspection have to work collectively so {that a} defect is caught on the step the place it happens, not after.
R&AN: Panasonic Join talks about “Gemba Course of Innovation”. How do you see automation and digital applied sciences altering day-to-day manufacturing operations on the manufacturing unit ground over the following 5 years?
SZ: Gemba is the precise place the place work occurs, the manufacturing unit ground itself. Moderately than staying in somebody’s head or buried on paper, the objective is to attach what’s occurring there with a digital system that may see it in actual time.
Over the following 5 years, I anticipate that to play out on SMT traces in two methods.
First, automation handles extra of the verification earlier than a changeover, checking feeders and settings towards the following job relatively than relying on an operator’s reminiscence.
Second, extra course of knowledge is captured on how expert operators truly work, in order that information can prepare new hires and tune the machines themselves.
Synthetic intelligence is being utilized throughout nearly each business. The place do you see AI delivering the best sensible worth in electronics manufacturing immediately, and the place do you suppose expectations could also be operating forward of actuality?
The best sensible worth immediately is predictive upkeep and defect detection to catch points utilizing machine knowledge earlier than they trigger downtime or scrap.
The place expectations run forward of actuality is the concept that a manufacturing unit will be run by AI alone, which is neither real looking nor the correct goal. AI is barely pretty much as good as the info it receives, and most vegetation nonetheless run legacy programs that don’t speak to one another, creating fragmented knowledge.
The actual alternative is to make use of AI to assist folks on the ground, taking up repetitive evaluation so operators can give attention to the advanced choices that require human experience.
R&AN: Producers are underneath stress to enhance high quality, enhance throughput, and stay resilient within the face of supply-chain disruptions. How can automation assist firms steadiness these competing priorities?
SZ: Automation’s actual contribution is consistency, for issues that are inclined to drift independently, like folks, machines, and supplies.
We give it some thought because the 5Ms: huMan, machine, materials, methodology, and measurement, as a result of points in high quality, throughput, and resilience sometimes hint again to a spot in a type of. That may very well be a employee underperforming, a machine drifting out of tolerance, or materials scarcity not caught in time.
Automated programs that monitor these 5 collectively can catch deviations earlier than they evolve into defects or missed deadlines and proper them in actual time. A line is resilient when one disrupted half doesn’t halt the remaining or ripple throughout it.
R&AN: Excessive-mix, low-volume manufacturing is changing into more and more widespread in lots of sectors. How are manufacturing programs evolving to deal with larger product variation with out sacrificing effectivity?
SZ: Excessive-mix manufacturing used to imply accepting slower changeovers in change for flexibility, although that’s altering. At this time, producers are decreasing that tradeoff by means of larger automation and course of intelligence.
Producers are investing in quicker, extra dependable setups, verifying supplies and feeder positions earlier than a run begins. On many traces, switching between configurations now occurs just by choosing a unique program relatively than by means of guide reconfiguration.
Feeder capability and early-warning programs that flag a part operating low matter simply as a lot because the software program driving the changeover.
Reliability is simply as essential as pace. Getting the changeover proper the primary time is very vital given workforce shortages and the rising problem of counting on skilled technicians who know the method by reminiscence alone.
R&AN: Panasonic Join has visibility throughout a variety of industries. Which sectors do you imagine are at the moment main the adoption of sensible manufacturing applied sciences, and what classes can different industries be taught from them?
SZ: There are a lot of industries, like automotive, client machine, and medical machine manufacturing, turning to sensible manufacturing applied sciences. Automotive, particularly, has led the shift towards constructing to demand as an alternative of holding stock.
That very same strategy is extending throughout electronics manufacturing, the place clients anticipate shorter lead occasions, so manufacturing has to reply rapidly relatively than draw from inventory.
Medical and aerospace add one other layer, since defects can carry expensive regulatory and security penalties, leaving no room for error.
The lesson for different industries is obvious: transfer past reactive inspection and towards real-time predictive monitoring, in order that deviations are captured as knowledge factors earlier than they turn into defects.
R&AN: Trying forward, what applied sciences or traits do you imagine may have the most important influence on electronics manufacturing over the following decade – whether or not that’s AI, robotics, digital twins, superior sensing, or one thing else completely?
SZ: The most important shift over the following decade might be convergence: AI, sensing, and automation feeding off the identical knowledge as an alternative of operating independently.
Many vegetation immediately have these capabilities, however are inclined to work in isolation, every system flagging its personal deviations with out the context wanted to grasp why they occurred.
What’s going to change is how these programs join. As an alternative of merely flagging a deviation, producers will be capable of hint it again to the basis trigger, whether or not it’s a worn half or a fabric lot, and use these insights to constantly optimize.
The factories that profit most might be people who evolve and get smarter from their very own manufacturing historical past, relatively than people who merely spend money on the latest know-how.
The aggressive benefit gained’t come from adopting extra instruments. It’s going to come from connecting them into a better manufacturing ecosystem.
