It seems that autonomous operations are finally within our grasp with the maturation of the various component technologies. But does that mean we are about to enter the autonomous age? Connected Technology Solutions spoke to Grant Coffin, technology consultant team lead of Rockwell Automation, for his views.
Is autonomous operation still an ambition?
I probably would not use the word ambition because it sounds like a destination, I would say it is the trajectory that we look like we are on, but I do not necessarily see it as an endpoint. As we get closer to new manufacturing technologies and new ways of working, manufacturing processes will undoubtedly develop. We need to bear in mind when we think about autonomous operations is why we are headed in that direction. It is underpinned by how organisations drive to improve value to the market and consumer through enhancing the quality. Even that environment is changing. You have got environmental impacts that are changing some of the dynamics around manufacturing technologies and processes. So, it is not just about the drive for lowering unit cost; there is a lot in the mix.
Would you still call it autonomous operations if you have a cell that can operate lights out as part of a more significant production line or bigger facilities?
I think it is part of the way there, and with autonomous operations in the extreme lights out form, there are lots of hurdles that we need to overcome to get there. There are many issues around different ERP integration from other organisations, which would be a pivotal element to provide insight into supply chains and the manufacturing capabilities of those supply chains and where they are at any one point.
People are adopting autonomous technologies when it makes sense from an operational point of view. If you look at where autonomous operations are used, it is predominantly in sectors looking to minimise human intervention for specific reasons. Let us think about Biopharma as an example. We do not want to be interfering with the environmental conditions within a cleanroom type environment where they are trying to maintain levels of humidity temperature within specific tolerances.
Most of my autonomous operation examples have been in the oil and gas sector, where they have some unique challenges regarding hazardous operating conditions. Do you think this scenario will remain the real prime driver, or will it be adopted in a traditional manufacturing facility?
I think you will see further developments in those areas. Still, I think there will also be an acceleration in other sectors, particularly areas of assembly and perhaps packaging within existing operations. If you look at some of the manufacturing technologies available now, such as independent cart technology that allows flexibility and agility, this has created a whole transformational environment around how manufacturing is executed. Those changes will continue.
What have we got now that that is enabling this? And what piece is still missing from the puzzle?
There is a congruence of elements that seem to be coming together now. The base enablement is the proliferation of computing power into the industrial environment, pushing a lot of edge computing and high-speed processing capability that probably did not exist before. That has been further enabled through the connectivity capabilities with the proliferation of Ethernet into the industrial environment. I would say open software platforms and unmodified Ethernet because that is going a step further to enable the convergence of the OT and IT environments. Bringing the enterprise together, traditionally, there is a lot of useful, high-value data within the OT environment that has never really been capitalised on in the enterprise space. The proliferation of Ethernet into the industrial environment is undoubtedly an enabler for that.
There is also the flexibility and agility enabled in manufacturing by new technologies such as smart sensing linked to other elements. This is where we open the conversation into digital twins, is if we think about our ability to do simulation and emulation now in augmented and virtual environments. This enables us to do elements of experimentation around processes mechanics and the control of those mechanics in a low-cost environment. We can try things in the virtual environment at a minimal cost that take a lot of risk and investment to achieve no output. That allows us to try new things, and our success rate when we deploy things down onto the shop floor is significantly increasing.
With technologies like Emulate 3D, we can link them to real control systems and, in a virtual environment, understand how they perform and operate as part of an overall process. Even to the point where we are introducing physical interaction, whether that be human or mechanic, into those environments as part of a broader system that allows us to experiment more.
It is not as simple as running these virtual simulations to find the optimum solution and then expecting it to populate all the plant’s control systems.
It is not, regardless of how good we can make them, and by the way, some of them are good. With Emulate 3D, the power behind it, for instance, is phenomenal in terms of its ability to work out the physics and everything that is in a real-world environment. But with the real-world environment, we cannot get away from it, it is different, and things operate differently, and things happen unexpectedly. To come back to where we started with the autonomous operation, I do not think we will ever really get to fully autonomous operations certainly in my lifetime. There are elements that humans bring, even from the point of view of creativity, to a process that the technologies that we have cannot do.
What other challenges are slowing the adoption of autonomous operations?
The specific challenges in the UK, and probably Western manufacturing, is that we have a cultural legacy. We have legacy systems, and there are many challenges around beginning the digital journey, and there are many organisations that are struggling with how they start their journey. New approaches are required; if you look at some of our acquisitions, adding to our internal capabilities, we have focused on bringing the domain expertise to assist customers in starting that journey.
Once people are on that digital journey and begin to automate processes, there are issues around the organisation’s culture and within the country itself. There is a significant skills gap in the UK and how the skill sets need to evolve are all massive challenges. Then there is the other end of it where people are concerned about what autonomous operations mean and the impact they will have from an economic and societal perspective to the point where we ask what the value of work is? What does it mean to society in terms of individual meaning? These are all challenges that governments need to be thinking about and about how they address them. We have seen the dynamics within the economic system today, and regardless of changes in technology over time, the market has adapted, and new skills have been required and developed. We would not have been talking about data scientists ten years ago; for example, the change in skills requirements are usually for higher-skilled jobs. I think the market will adapt over time, but these are all challenges that need to be addressed.
This goes back to where I began. It is not primarily about reducing costs, although everybody sees it as we want to minimise cost and make things as cheap as possible. But it is more about the value that an organisation brings now because that is what people are interested in.
Can you talk to me about anything you have worked on?
If we break it down to a cellular level, we have been doing autonomous operations for several years. I started with Michelin in Dundee in 1987, so I have been in manufacturing and living the changes over my 12 to 15 years on the shop floor, and the changes were enormous. Those changes are continuing, and one of my job’s joys is I get to see them across industry sectors. In terms of specifics, some of the newer capabilities that we have, like the ability to improve significantly data acquisition and optimisation, have significantly changed how we would class it; it is not just a robot in a cell anymore.
We have had things like Model Predictive Control for a long time, but this access to large data sets means doing things like analytics and taking it much further. Organisations are doing that to create a competitive advantage by improving their sustainability holistically.
Where do you think we might be going in 10 years?
I think society will drive how far do we want to push things like autonomy? We touched on the fears earlier on. There are elements of trust and societal trust around what autonomous operations mean and how that may expose us to some of the threats we perceive to be out there now from a government and economic position. I think it is how governments approach it as well, in terms of how you look after the social aspects, that they need to look after. To sum up, I am not 100 per cent sure we will see the market change with more autonomous operations and more opportunities for jobs with higher skills requirements, some that we probably have not even dreamt of now. It is by no means doom and gloom; I think it is an exciting time for automation. It will create new opportunities, expand our possibilities, drive our sustainability and ultimately, our ability to explore the universe further.