Can we turn off
the lights in factories?
“Lights out” manufacturing is used for manufacturing processes where factories run autonomously without human intervention. The term is quite literal, with production occurring free of human necessities like lighting or heating, ventilation, and air conditioning (HVAC). Famous Sci-Fi novelist Philip K.Dick first referred to it in his “Autofac” short story in 1955.
There has long been a promise of a smart, everything-can-do future. On the other hand, it’s hard for many manufacturers to look around an existing facility and imagine that it could one day be transformed into a factory of the future – whether lights-out or otherwise data-intensive to reach new production efficiencies and quality of output.
But, with the advent of Artificial Intelligence and the digitalization of the shopfloor, this is likely to make its way from fiction to reality in the coming years. However, although an increasing number of steps and manufacturing processes have become fully automated in many industrial sectors, the way to turn the lights off is still full of challenges and pitfalls.
The number one technological cornerstone in reaching this stage of complete production automation revolves around data. Data is everywhere in the factory: from CAD/CAM, PLM, and simulation systems to sensors at the end of the manufacturing line. But much of this data is unstructured. To turn big data into actionable smart data, it must be aggregated and contextualized. This structure connects data points with their meaning so that digital tools can be used to
detect trends and identify problem areas. Insights like areas of high downtime for a specific machine, too many operator errors, product build or supply issues, long lead times due to monotonous, operator-intensive work, or an overabundance of non-value-add activities become visible when manufacturing data is harnessed, organized, and evaluated.
Data is not being used to its full potential but in isolation, though. In many factories, various departments work on systems that are not interconnected. This means, for example, that a quality issue detected in a work-in-progress (WIP) on the factory floor is difficult to connect to a variation in the properties of an incoming component or raw material that may have caused the WIP problem.
So, the journey to lights-out and other high-efficiency production schemes begins with something already present in manufacturing operations: data. Connecting it across systems and with meaningful concepts often takes the form of a digital twin: an interactive, virtual representation of a physical product or process. Digital twins enable engineers to explore alternative approaches and what-if scenarios before they take the plunge with significant capital investments. Digital manufacturing tools and the digital twin are designed to replicate autonomous processes in the virtual realm, where they can be optimized before deployment.
In a nutshell, the technologies already exist to further automate the manufacturing of goods, but are we ready as a society?
As with every such digital transformation movement, the power of tools is not the only nor the most critical success factor of change. Cultural change inside companies and broader society is far more essential and is two-sided.—On the one hand, automation will replace current production jobs and less qualified workers, threatening our social classes with displacement and our societies with unrest.—On the other hand, adopting such lights-out factories will create the need for new jobs around manufacturing. A whole family of new positions and skills is likely to emerge in the design, planning, control, monitoring, and maintenance of such autonomous production sites.
How will companies and public powers accompany this shift towards a new economic organization? As in many cases, technology will succeed at scale if it is perceived as a tool for humans, not if it results in a threat to them.