Automation software comes in different clothing and performs different tasks. It could be a programming integrated development environment (IDE) such as RSLogix5000 or CodeSys 3.x, or a SCADA application created in Inductive Automation’s Ignition. It all comes in different flavors and colors.
Artificial intelligence (AI) is the new kid on the block and is coming on very strong in certain sandboxes.
Machine learning applications are not new, but the methods used are. Is AI ready for prime time? It’s sure getting there.
Sharmeer Mirza, who is a senior research engineer at PepsiCo, recently gave a talk on how they are using AI and machine learning to create a better potato chip. It is based on sensing the texture of the chip without harming the little fellow.
The premise is to bounce a laser of the chip, and correlate the returning sound(s) to texture creating a quality check for the product.
If this sounds like adaptive process control (APC), I would say you are right. But using end product statistics along with APC to create a better chip may be considered novel.
Machine learning extends into various industries such as biology, blood systems, neural networks and the like. Most would argue that to be AI you have to use an algorithm, but in this early stage in AI machine learning the jury may still be out.
Adaptive learning may be the automation software that leap frogs us into true AI applications.
Also we need to remember that the control engineering discipline has yet to catch up to the AI learning curve.
My good friend Dick Morley, father of the PLC, told me many times that point-of-sale manufacturing will come to a Home Depot near you. There are already many instances of mobile manufacturing, such as pizzas and kitchen cabinets.
But what about a machine that can create a single unit of Product X and then a single unit of Product Y? That’s adaptability and flexibility.
CNC machines of current day with an automatic tool changer isn’t AI but is adaptive based on its g-code and m-code. A 3D printer can also be considered an adaptive machine, but then most printers are.
Single-use devices such as a rotary phone, as we know, are dead in the water. Can multi-function be described as adaptive?
B&R Automation, along with Control Design, wrote a white paper on adaptive systems instead of having dedicated machine designs. Hmm, one might say!
Automation software wears many hats as suggested. So what member of the software community allows for a batch-of-one mentality?
John Kowal, B&R director of business development, suggests that each product can be individually manufactured as ordered, like McDonald’s. You order your product; it gets made and delivered. No finished goods inventory control required. In this day and age of trends and fads, investment in manufacturing can be made knowing that the new-age production machines will be re-purposed after the fad wanes in popularity.
Kowal makes an interesting observation, something he calls, “digital native.” This is where a majority of youth have grown up with the digital age of mobility, access and information everywhere. They expect things and, along with the McDonald’s mentality, want it now; manufacturing must adapt to keep the masses connected to the buying process.
IIoT is not legacy, but it will have to integrate with legacy somewhere. This will take both hardware and automation software, perhaps yet to be developed, but it will come.
Inductive Automation’s Ignition is a prime example of how flexible an HMI/SCADA system can be, as well as supporting the new paradigm. The way it integrates new technology and not have things such as MQTT as a bolt-on lends itself to the tasks at hand such as cloud data easily.
The Industrial Internet Consortium (IIC) was created to foster the development and adoption of interconnected devices and machines along with the people who develop and use the technology and software. The IIC runs an IT/OT task group, which has proclaimed that our new age of networking with devices is not about having these two work independently, but to create the culture that allows them to work together to solve the issues that will invariably raise their ugly heads with the new order.
It comes down to standards. There are many bodies that are developing certain standards, and there will be many companies who will engage these standards in order to gain market share and to do the right things.
Make no mistake, it will be the software that will drive the integration and sense of community in our future world.