Industry 4.0 vs. IA2AI From Automation to Autonomy

 

Industrial Automation to Industrial Autonomy(IA2IA) is a relatively new term. (see InTech May/June 2020, Autonomy and beyond by Dr. Ted Abe https://www.isa.org/intech-home/2020/may-june/columns/to-autonomy-and-beyond ).

source: https://www.yokogawa.com/special/ia2ia/definition/

The long term goal in many industries is to transition from automation systems into increasingly autonomous capabilities. The concept is not new. Scientists and engineers have envisioned flexible automation at the early beginning of robotics. Today, we are taking the idea to the next level, automating tasks that are not necessarily repetitive or routine needs a solution that requires input from several areas of artificial intelligence. Knowledge representation or planning systems are areas that have been overshadowed by the excitement of deep learning.

Nevertheless, new architectures of artificial neural networks are not a magical solution for all problems. Such tasks as ordering a sequence of discrete steps to reach a goal state (finished product) are not a trivial problem. Anybody who studied the "Sussmann anomaly" should be able to see the potential issues. (https://en.wikipedia.org/wiki/Sussman_anomaly)

On the other hand, machine learning or deep learning as a tool for programming robotic/automation systems appears to be in the future of industrial systems. We see the first signs of this evolution in discussions about new SCADA and other design or supervisory systems. Transitioning a production line from one product to another is always a costly task. Data-driven automation, in which the system learns how to put together a sequence of steps, will revolutionize manufacturing.

https://www.oilreviewmiddleeast.com/industry/oil-and-gas-chemicals-petrochemicals-companies-to-adopt-full-industrial-autonomy-by-2030

Another area of critical importance is the hardware configuration. Cyber-physical systems need to be physically reconfigured to achieve genuine autonomy. This concept has been at the forefront of development in the area of industrial machining. Many vendors strive to take flexibility to the next level in their additive and subtractive machining systems. Perhaps, the IA2IA revolution will start on a small scale inside the machining system that contains all the tooling and software to support autonomy. We will soon realize that a single machining system could complete many diverse tasks previously done by a multi-stage production line equipped with several specialized tools.   

 Autonomy is a factor that contributes to increased flexibility when dealing with environments that can not be modeled by simple feedback loops. The cost of human labor has always been critical and will continue to grow. New technologies like nanotech will put increasing pressure on organizations to remove human labor from the production/operations process.