Smart Factories – The Future Of Manufacturing

COVID-19 and the Suez Canal debacle reinforced one topic for business leaders across the world i.e. you can’t have a single source for all your manufacturing. While you can’t compete with China on pricing, what you can do is automate. We aren’t simply talking about robots or adding two more members to your IT team, but Smart Factories. Smart Factories are fully digitalized factories that use sensors to constantly monitor and share data that helps bring more efficiency and predictability to the manufacturing process.

According to Wikipedia, Industrial Revolutions have traditionally been the transition to new forms of manufacturing that increased the output by a significant amount. We have had 3 individual industrial revolutions so far; the first was powered by the steam engine which helped us harness the power of machines, the second was the assembly line that helped us mass-produce more products (read cars), and the third was the availability of computers and automation which helped us move repetitive, low-value tasks to specialized machines. We are now on the cusp of the 4^th Industrial Revolution which is expected to be powered by data and knowledge. 

With the advance of technology and automation, everything around us is getting smarter, faster, and more customizable to our immediate needs. Similar to how fitness devices track our movements to recommend actions, several processes need tracking and improvement on a factory floor. In its simplest form, you can imagine a smart factory to be a factory floor where most of the work is done by machines, and each step of the process is tracked, measured, and the data is relayed to a central data warehouse. These sensors are part of the Industrial Internet of Things (IIOT) that constantly track and measure each process so that there is more predictability and knowledge of points of failure, and sends it all to the central data warehouse (could be cloud). In addition to improvements in design and speed, smart factories help tackle three major issues that affect manufacturing: 

1. Maintenance & Downtime: Downtime in factories is when the machines aren’t running at their optimum because of part failures or regular maintenance of some portion of the machinery. Usually small, these failures can sometimes grow to cripple the entire production line if spare parts are not found in time. Predictive maintenance is the major selling point of smart factories where workers can anticipate issues and order the necessary spares in advance and schedule repairs.
2. Lack of Data for Advanced Modelling: Even though modern factories have evolved from the clip pads, the amount and quality of data isn’t sufficient to support the adoption of AI and Machine Learning (ML). AI & ML can help identify patterns and opportunities for improvement across the supply chain and support efficient warehouse management.
3. Lack of Skilled Labor Onsite: Smart factories use Digital Twins, which involve computerized models of processes and equipment to troubleshoot from afar. These are different from simple computerized models as they rely on real-time data. This not only helps reduce costs but also helps bring expertise to an issue from around the world. This concept is most famously used by NASA to repair spacecraft and the International Space Station (ISS). 

In addition to fixing these issues, smart factories also provide a significant advantage by allowing companies to place their factories anywhere around the world without worrying about the availability of a large blue-collar workforce. With all these benefits, there hasn’t been a complete implementation of a smart factory so far, but it is not due to a lack of trying. Some of the limiting factors that have temporarily prevented smart factories from being realized are: 

1. 5G or Similar Technology: According to the Thales Group, the current capability of 4G can support only up to 4000 devices per square kilometer, whereas 5G can support up to a million devices in the same area. This is a major limiting factor when you introduce all those IIoT sensors onto the factory floor.
2. Battery Life: With all those sensors on the factory floor, it would be brutal to go ahead and replace the batteries on all of them. 5G is projected to improve the battery life by up to 10 years.
3. Low-Quality Master Data: To fully utilize the data generated in the factory, companies need a single view of their ERP data, customer data, and IIoT data. Issues with their existing data and competing priorities have prevented companies from embarking on a mission to clean up their data.
4. Advances in Edge Computing: With such a large volume of data being generated, it would be beneficial to parse through and process it on the sensor location as opposed to being uploaded to the data warehouse for processing. With current computing technology, it could be very expensive to upload and process this data and may overload the network. 

Although these limitations exist, they will only be a temporary impediment to progress in the Smart Factory story. Even though full-fledged smart factories may not spring out of the ground in a year, companies will begin to “smartify” certain parts of the factory floor and analyze their benefits. This will be a critical success factor to justify the cost of building out a smart factory and if it indeed lives up to the hype. As an end consumer, how this impacts you is when you buy directly from a manufacturer and they can accurately guarantee the date of delivery. Those who tried to buy cars during the pandemic or if you were trying to get the latest iPad delivered to you, you know the pain when you were told delivery in 4-6 weeks! With the data-driven predictability and real-time view of the factory floor, companies could have a better insight into product readiness and availability.