Focus on Lights-out Processes, Not Lights-out Factories

SC_B 201366 D200714 July Cover

Factories are volatile environments. Varying utilization rates, staffing levels and staggered shifts — and the new demand to create a consistently safe work environment — each have their own challenges that require attention so demand can be reliably serviced.

Meanwhile, it is clear and encouraging that no one is hitting the pause button on their smart manufacturing plans anytime soon. Investments are, however, being pivoted toward the worker experience. This is evidenced through client inquiry, my work with industry associations like CESMII and the practitioners that participated in recent research on return-to-work strategies for factories that my colleague Dana Stiffler and I just completed.

Is there any insinuation or discounting discussions on factory automation and “lights out” manufacturing? Those dialogues have only intensified. Interest in “lights out” manufacturing where factories can run with extremely minimal intervention has reappeared. The concept is not new. In 1982, General Motors embarked on a lights-out journey at a plant in Saginaw, Michigan. The objective? Gain competitive leverage by increasing productivity through automation. Even though the site achieved a sub 10-minute changeover time from one axle design to another, it was eventually closed as part of a cost-cutting exercise in the early-1990s.

Fast forward to 2020. Are we any closer to lights out than we were six months ago? No.

Even before the pandemic, promises of fully automated, hyper adaptable production were suspect. Only 17% of the respondents in our 2019 Realization of Industry 4.0 Study believe they will run completely on digital, lights-out production setups by 2025. Meanwhile, three-quarters (79%) see their manufacturing operations as being composed of human-driven, manual processes augmented with digital ones. On a more cynical note, let’s face it: for some companies, removing paper or eliminating a manual process would be a revelation.

SC_B 201366 D200714 July Graphic1
Let’s not misinterpret the data. Automation is a necessity in manufacturing and there are ample opportunities. Yet when combined with the current state of affairs, a judicious strategy versus a “reactive” or “defensive” approach is sensible for the long term. Over-automation can be disastrous. It was the culprit behind Tesla’s falling short of its audacious goal of 5,000 vehicles per week. Elon Musk later copped that “humans are underrated.”

Trade-offs still need to be made. Labor versus automation trade-offs persist and aren’t just a discussion for economies with high labor cost or dwindling factory skill sets. In addition, ongoing reshoring and location shifts must be factored in.

Some things have not changed. Automation is still menacing to factory workers. However, it’s not just the physical form factor of a robot or automated guided vehicle that will remove and replace some roles and skills. Today, it’s varying forms of artificial intelligence (AI). Computer vision will lessen the need for close inspections, while robotic process automation (RPA) promises to eliminate the busy work for engineering and quality control roles. The returns are attractive. Imagine reducing inspection processes by 90 seconds, with ongoing improvements in accuracy.

While some of these investments can be perceived as “improving the worker experience” and are sold as the “elimination of non-value add,” we need to look at the long game. Will today’s liberation of a worker to “do more with less” inadvertently create tomorrow’s overhead reduction target?

Today’s manufacturing environment makes automation even more personal. Similar to removing restrictions and coming out of quarantine, automation investments need to be tactical and gradual. Balance is necessary. It means suppressing “knee jerk” or reflexive investments and diligently identifying where your opportunities are. We cannot control our reflexes, but we can control the ability to continually adapt toward an uncertain future by approaching factory automation this way:

  • Focus on lights-out processes, not lights-out factories.The supply chain’s demand for factory flexibility and speed will increase as activity picks back up. Where can processes be at least 75% lights out? Managing the movement of materials becomes a ripe target. Eliminating dwell time between stages of production or shortening the time to bring materials to production lines improves production rates and positively impacts controlling inventory levels and supplier interactions.
  • Right-size the approach.Where can the operator’s experience be improved with a digital tool? Immersive experiences (AR or VR) provide a “coach and advisor” capability that marries data and intuition. This creates consistency and shares knowledge — versus removing it — without compromising certain methods and procedures. Changeovers are ripe for this approach.
  • Don’t react.Today’s “non-value add” is tomorrow’s “inadvertent overhead.” Will automating a set of tasks to lessen the burden on a worker today create tomorrow’s overhead reduction target?
  • Prepare for what lurks in the financial shadows. Factories are far from updating technology with the same frequency as an application on your phone. Technical debt is consistently overlooked and integration costs are always (significantly) underestimated. Be prepared to spend money modernizing operations technology (programmable logic controllers, drives and other technologies that control a process).

Automation is a serious topic and will reshape the way factories function and how manufacturers compete with different products and services on global and local basis. In the spirit of levity in these unique times, I’m reminded of this quote from Warren Bennis:

“The factory of the future will have only two employees, a man and a dog. The man will be there to feed the dog. The dog will be there to keep the man from touching the equipment.”

Don’t turn the lights out yet.

Simon Jacobson,
VP Analyst,
Gartner Supply Chain
[email protected]



Author Simon Jacobson

More posts by Simon Jacobson