The future of automation in manufacturing

The next few decades will be an exciting time for the manufacturing industry as the pace of adoption for automation increases. In this article we’ll look at some key technologies that will build the foundation for the future of automation in manufacturing.

Artificial Intelligence

Artificial Intelligence (AI) is often used as a buzzword to make things sound exciting and interesting, however, the AIs often reported are simply machine learning algorithms. True artificial intelligence is anything from 5 to 50 years away, depending on who is being asked. The true benefit of artificial intelligence in a manufacturing set-up would be vast. Imagine a system that knows every single thing about the robots in a factory; a system that can make business decisions based on predicted raw material prices and manage the factory as close to its theoretical maximum productivity as possible.

Once the first AI is developed the entire face of all industries will change in ways that cannot even be predicted, however, one thing is for sure, productivity and efficiency will reach levels previously thought impossible. The many unknowns faced when speaking of AI mean that its benefits cannot be seriously considered without making wild assumptions. More realistically, machine learning algorithms that are currently being used in industry and are continuously being improved will have numerous benefits in an automated factory. These are listed below.

  • Supply chain management – Required stock levels can be determined by analysing demand forecasts. This will ensure that there are always enough raw materials to manufacture products.
  • Logistics – Returns can be automatically logged, evaluated and actioned with minimal human input. Furthermore, dispatch and arrival times can be accurately calculated based on historical data. Predictive analysis can also be used to an extent.
  • Predictive maintenance – Embedded sensors, as well as general machine monitoring, can be used to decide when a machine is about to break down. For example, the vibration of a machine can be monitored and if any anomalous behaviour is detected that vibration profile can be compared to a library of known failure modes and an appropriate action chosen.
  • Data management and analyses – With all the sensors and monitoring going on in the automated factory of the future, there will be a wealth of data created that can be turned into actionable information that a factory manager can then use to justify necessary changes or upgrades. More radically, the data can be gathered, interpreted and actioned on with no input from any human. This opens the door for a factory to autoregulate its operations to reach its predefined goals.

Collaborative robots

Typical industrial robots are blind and can seriously injure any human who gets too close; this is why robots are typically enclosed behind fences. However, a new class of robots known as collaborative robots or cobots has emerged that can operate safely around humans without the risk of injury. Cobots are designed to work alongside humans on the same tasks; these systems need advanced sensors and software systems to accurately predict and work with a human operator. Human-robot symbiosis in manufacturing plants is 85% more productive than either element alone.

Another key advantage of cobots is their ease of programming. Traditional robots required highly-specialised programmers to effectively set them up. Cobots allow human operators to teach them to do a task. This teaching can be as simple as grabbing the robot and moving it manually to mimic an action. This allows workers to train their very own little helper who won’t complain or get tired. 

Lights-out manufacturing

Despite the previous section’s focus on human-robot symbiosis, some factories can run more efficiently without human interaction. This is the case for plants in which high rates of repetitive manufacturing take place or where the range of products manufactured is relatively small. 

These factories can theoretically be completely automated from when the raw materials are dropped off in the warehouse to them being stored as inventory, taken to CNC machines and set up. Thereafter, the manufacturing can take place and the part can be removed and taken to an area for inspection and subsequent packaging. This entire process can happen with no human operator in sight.

Factories that are capable of 100% automation will be the gold standard of the manufacturing industry as they will generate impressive profits after all the machinery is paid off. Obviously, there will be breakdowns and unforeseen issues, and realistically a human supervisor will be required to ensure any unexpected event does not damage the highly expensive equipment.

In conclusion

The natural progression of automation in manufacturing is always towards more automation. In the not too distant future, fully automating a factory will not only become technically feasible but also economically prudent. In terms of the hardware required for full automation in manufacturing, current robotic systems are extremely capable, however, the software ecosystems that need to be developed and put in place to allow for a fully automated process are not quite there yet. This is why the future of automation in manufacturing will require a combination of advances in software and hardware. Refer to our automation whitepaper for more information on how automation can benefit your company.