Crises never stop visiting the human civilization, to interrupt the normal course of routine and to expose the fragility of our plans. Apocalyptic situations like the COVID-19 pandemic can bring a large part of the world to a halt and can bring the economy down to its knees, begging for a respite. Such situations lead to complete erosion of workforce from the offices and workplaces. And though employees in IT services can continue working on their laptops, the people working in factories and production units face a huge challenge as they cannot visit the production site for daily operations. This ails the economic output and makes people wonder if economic production should always be subject to the games played by nature. Should we make peace with the idea of submitting to such catastrophes and stop production, which leads to losses in millions and impacts the country’s Gross Domestic Product in billions? Or should we take a step to keep factories running, without risking the life of people?
No wonder which idea sounds good. Companies around the world have realized the need of agility in production and the need to make factories run on their own. Corporations are exploring ways to make factories think by themselves, plan by themselves and be able to fend for themselves in times of disasters. It is the era to make factories intelligent and make them autonomous in operations. This is the era of the Autonomous Industries. The entire fleet of cutting-edge technologies, ranging from edge analytics, remote monitoring, drones, artificial intelligence and augmented reality, will infuse life into the Autonomous Factories.
Predictive Analytics to keep factory alive
The factories of future will have IoT sensors in all the machines and equipment, which will keep monitoring the health of the machines 24/7. The sensors will keep a tab on the functioning of machines, flow of fluids, transfer of good from one location to another and rigorous quality checks. Any anomaly in the normal functioning of the machines will result in an alert that will be sent straight to the central data center. This central hub will be responsible for interpreting the error logs generated by a machine or a group of machines. Predictive maintenance algorithms will run over the error logs and the central command will figure out the best procedure for addressing the technical issue. Bots will be assigned to attend to the failing machine and these mechanical bots or software bots will bring the machines back to life with the help of all the procedures known to the repair manual repository.
However, if by any chance the bots fail to get the machine in its desired condition, then the machine will be detached from the production chain and the entire production system will re-optimize itself among the other healthy functioning machines. In the meantime, the error logs along with output of all repair processes will be sent to a production site manager, who will be controlling the factory remotely. The manager will then deploy a few technicians to look into the machine and to bring it to life.
Edge Computing to accelerate intelligence
When machines generate performance data related to the inputs received, output generated, temperature levels, climate conditions inside the factory and other vital parameters, it is certainly a lot of data generated every minute. Uploading all of that data on to cloud, next perform computing on cloud and then further relay back the insights to machines takes a long time as the volume of data increases. However, sometimes analytical results are required at an instant, which is where such latencies can prove to be fatal to the supply chain. As a result, analytics needs to be done closer to the machines, i.e. the source of data.
Edge intelligence will solve such issues of latency in data analytics and insights generation. It will bring computation power close to the machines and will reduce the amount of data uploaded on cloud. The data generated per second will be filtered off the unnecessary pieces of content and only the useful information will be relayed for further deeper analytics. It is just like our nervous system. The brain does all the processing, but when intelligence is required instantly for a reflex action, spinal cord takes the charge to accelerate the speed of computing. That’s exactly what edge computing will do to make factories run on instant analytics so that any kind of anomaly is identified and worked on at the earliest.
Remote Monitoring using Drones
Autonomous flying machines – the drones – will help keep a watch on the entire production unit and they will leverage computer vision to understand the environment around them. Drones will be instrumental in inspecting the entire plant and also each individual machine. In case any damage is detected in places that are not connected through the IoT sensors, it will be recorded by the drones and the information will again be shared with central command. The command center will chalk out the best plan to resolve the issue and will either deploy bots into action or will inform the human manager to get on his toes.
Drones will also be used to inspect the condition of the outer structure of the factory, the impact of weather on the structure and the areas that need mending through labor. Identifying decaying areas in advance will help to fix the issues proactively, which will save from big losses or expenses later and will keep the factory running without disruptions.
Demand Forecasting and Supply Management
Disruptions in supply chain are inevitable when crises like pandemics, or natural disasters occur. This is when the production planning managers are faced with one of the arduous tasks of calculating how the demand will fluctuate and how the supply must be altered to keep losses low and inventory optimum. However, machine learning models may soon pick up this task from humans and will be capable of forecasting the optimum mix of raw material purchase for safety against uncertain future supply and production of goods to cater to uncertain demands in the market.
During an impending natural disaster like floods, a water bottling plant will forecast that the demand for pure drinking water will go up. Whereas a toy manufacturing unit will understand that it is going to see a slump in demand for a few months. However, it is not sure if the raw materials for bottling will be available going forward. Therefore, the water bottling plant will try to ascertain the amount of raw material it must procure to safeguard itself against disruption in supply of raw materials. Similarly, toy manufacturing factory will lower its procurement of raw material. These decisions will be autonomously taken by the central command in the intelligent factories, which will lead to higher autonomous agility in the production process. Based on insights of regression models working on historical data, factories will autonomously handle procurement, manufacturing and inventory for keeping itself valuable and sustainable during and after the crisis.
Augmented control over processes
The production site manager, in charge of overseeing the entire production facility will not feel isolated while being far away from the production site. Though the production facility will be functioning autonomously, the manager will be very much present in the factory, through enhanced reality. Virtual reality headsets will take the manager into a virtual realm which will simulate the interior of the factory and will give the manager a deep look into the machines, the output of final goods and the entire autonomous factory. Whenever the manager will need to know more about a particular machine or a process, she will simply tap on a virtual button placed near the machine and an augmented reality-enabled pane will open up consisting all details of the machine. The vital parameters of the machines will be flashing right in front of the eyes of the manager in that virtual world. Using enhanced reality systems, the manager will have complete control over the factory that on the outside looks totally autonomous.
Conclusion
The factories of tomorrow are going to be much different and much advanced that the factories of today. The factories of future will be agile, autonomous and adroit in managing their own affairs, while humans will also feel being in much greater control of the processes. Humans will no longer be involved in mundane labor work, but rather in the more intellectual aspect of leadership. This is the future of the factories which is not much far away from us.
