Should Lights-out factories be the standard
for every manufacturer?
Over the past few months, a slew of lockdown measures was implemented to contain COVID-19. The blow from the pandemic and lockdowns have destabilised economies and impacted many industries. Mobility has also been greatly restricted, forcing people to change their lifestyle and the way they worked—rather than commuting daily to work, remote working has become the ‘new normal’. While this transition has been relatively smooth for some, it has been a challenge for manufacturers who, by nature of the industry, relied heavily on manual inputs by operators.
As labour becomes a scarcity, manufacturers find themselves having a renewed interest in digitalisation and automation. Even as countries are partially reopening and manufacturers are easing into the new normal, anticipation in the area of ‘lights-out’ factories remains at an all-time high.
What is a lights-out factory?
‘Lights-out factory’ is a factory that runs autonomously with minimal or no human intervention. By relying on intelligent machines, Artificial Intelligence (AI), big data and interconnected digital communications, the factory is free from manual labour and it can operate without the usual “necessities for human operators” such as heating, air-conditioning, or lights—hence the term ‘lights-out’.
To successfully implement a ‘lights-out’ factory, there has to be more than just automation. Effective quality control, planning, maintenance, and logistics are integral to ensure that the machine can operate round the clock.
Is a lights-out factory right for you?
1. Reduced risk of contamination
Theoretically speaking, a fully autonomous operation is particularly useful for manufacturers who are highly susceptible to contamination, such as those in the food, biotech, nutritional products or microchips industry.
Fig 1. Food hygiene is of utmost importance to food manufacturers. Mishandling or contamination due to human factors can prove fatal to consumers consuming the food.
Contaminations are not only costly to clean-up and recover, but they can also lead to serious safety and quality issues. Without human interactions, however, the risk of contamination may be greatly reduced.
Higher precision while working round the clock
“Machines do not grow tired. Once programmed to work, they can repetitively churn out output after output with little degradation in the quality.
Unlike human operators who are easily fatigued and cannot work 24/7, machines can work more consistently with significantly lesser ‘rest time’ (i.e. downtime). Machines are also capable of higher accuracy and levels of precision, be it when performing delicate tasks or when doing repetitive work.
With the relevant sensor technologies and software integrated, machine data is constantly collected, measured, and analysed to make sure that outputs are produced according to the highest quality. Should there be deviations in process parameters, the software would detect them. Ideally, AI and machine learning (ML) technologies would then automatically reconfigure the manufacturing environment and resolve the problem. Every successful implementation would render the technologies smarter and improve their accuracy and speed to correct issues more promptly.
Reservations towards ‘Lights-out’ factories
Despite the excitement around ‘lights-out’, we need to remember that maturity in the area of autonomous technologies has not been achieved. Most of these technologies are still in the growth, if not the introduction stage. It is therefore too soon to conclude that manufacturers can rely solely on automation to produce and foster long-term growth.
Moreover, full-scale automation can prove more costly than expected
In an autonomous production environment, one would expect the level of labour and capital efficiency achieved to be much higher as compared to the traditional production model involving human operators. And of course, with greater efficiency comes lower cost of production and higher profit margins.
However, before rejoicing in the cost-savings received, let us not forget this: manufacturers cannot experience the efficacy and cost-savings from automation without first making significant CAPEX, and high CAPEX can be extremely risky. It is not uncommon for manufacturers to fail to receive their expected returns on investment despite the heavy CAPEX made. For this reason, manufacturers must consider the benefits and cost-savings achieved from full automation and weigh them against the risks involved with the necessary CAPEX to be made.
There are also many instances where semi-automation are masked as ‘lights-out’
Manufacturers tend to think that the fewer the number of operators on the production floor, the faster they are inching towards ‘lights-out’—but this is not true. As machines and automation free up operators and managers from rote tasks, they proceed to make more strategic and analytical contributions behind-the-scenes. As such, the factory is not running autonomously because human intelligence continues to intervene in the production process (in the background).
For example, while a factory may be using Automated Storage Systems, Automated Guided Vehicles (AGVs) and Autonomous Intelligent Vehicles (AIVs) to automate the transportation of internal logistics, operators are using tools such as real-time factory-level dashboard to decide and direct how and what materials should be transported between the warehouse and factory.
Furthermore, when there are exceptional errors or unforeseeable problems such as the sudden breakdown of machines, the maintenance managers would have to manually intervene and arrange for corrective maintenance to be performed.
Fig 2. Operator adjusting the valves of the hot water machine so the beverage processing process can run smoothly
“In reality, ‘lights-out’ factories remain an exception and are still not widely prevalent among manufacturers. There are even fewer discrete manufacturers—those producing units with high complexity and low volume—who are operating in a ‘lights-out’ manner.
Discrete manufacturers such as those producing vehicles, aircraft, smartphones, computers and clothing, by their complex nature, require special dexterity which humans excel at. This shows that while there have been great advances in autonomous technologies challenging humans, human intelligence and labour continue to be highly valued.
Promoting collaboration between humans and technologies
Rather than perceiving humans and technologies as ‘mutually exclusive’, manufacturers need to consider them as ‘mutually beneficial’.
“More attention should be placed on complementing humans with machines and automation, and not to replace humans with them. Use technologies to remove redundant tasks, help operators do more with less, and transform the manufacturing process.
From the first to the fourth industrial revolution which we are currently in, technology has always been an enabler and extension of humans, but not a replacement. For these reasons, lean manufacturing and digital transformation concepts—both of which focus on using technologies to help humans achieve greater efficiency and better business results—are extremely popular and widely adopted by manufacturers globally.
Fig 3. An aeroplane factory in the 1940s. Human operators and machines working together.
When technologies and humans work in harmony, manufacturers can expect prompt actions taken based on real-time production data.
By connecting machines on the production floor with IIoT sensors, real-time data is collected directly from the machine and sent to the integrated data analytics platform to measure metrics such as Overall Equipment Effectiveness (OEE) of each production line, cycle time, as well as to monitor the health and status of each machine. With these valuable data in the bag, manufacturers can proceed to:
Improve OEE. If a manufacturer aims to increase OEE, they can simply click into the analytics software and see the areas that are affecting Availability, Performance and Quality. For instance, if a bottleneck is causing extensive downtimes, production managers could immediately identify where it is and promptly take actions to eliminate it.
Analyse and resolve anomalies. If the engineers would like to look for anomalies and resolve unusual behaviour
before they affect the performance of the factory, they can review and compare the data period by period (the last hour, the last day, the last week, the last month, etc.) to have a better understanding of the problem. Should the anomaly require attention from other departments or stakeholders, they can even be flagged out for further investigation.
Visualise production floor performance. Real-time production floor data can be displayed for not just the C-suite executives to have an overall view of the factory performance, but also provide operators with opportunities to visualise their production performance and catch-up if they are falling behind their counterparts.
‘Lights-out’ factories and autonomous technologies are undeniably generating value for manufacturers and some sectors use them more than others to reduce the risk of contamination and cope with labour-crunch issues. However, these technologies are mostly still in the development phase and they do not capture as much value as humans and technologies combined. While we work towards the ‘lights-out’ factory model which is about more automation and smart manufacturing, it is essential to remember that human capital remains highly valuable. Rather than replacing one with another, manufacturers should look into the potential value that will arise from humans and technologies coming together. We have outlined how there would be improvements to the way workers engage with their jobs and greater real-time connectivity throughout the factory, but really, the possibilities could be endless.