The Eye of Oden podcast – Episode 3: Machinery veteran Steve Braig on the future of automation and robotics

In episode 3, Steve Braig, Oden’s VP Business Development talks to us about the the future of automation and robotics.

VO: Welcome to The Eye of Oden. The stories behind the team that brought you intelligent industrial automation and developed the industry’s first machine learning and artificial intelligence for manufacturing. The team who’s working tirelessly to deliver new technologies to the factory floor, and to help manufacturers optimize their production, eliminate waste and close the skills gap forever.

Let us introduce you to Steve Braig, Oden’s VP of Business Development responsible for building relationships with manufacturing equipment makers and strategic partners.

VO: Steve is a self-professed manufacturing lifer, but what attracted him to this industry in the first place. He gives a few reasons.

Steve: The first one probably always has been in my DNA. My father, my grandfather, my uncle, my brother – all engineers and back at the time it was just a natural path for me to follow. Secondly. I’m always excited about seeing a product being made, seeing physics at work. A fast moving toggle clamp on an injection molding machine or precise motion controlled by server technology. It excites me. It gives me a lot of satisfaction. Lastly, we all want to do something in our life or in our career that is meaningful and contribute to society. Manufacturing has the largest multiplier effect in any industry. So for every manufacturing job that is created, there is about one and a half jobs in support of that particular activity. So the socio-economic benefit in manufacturing is far larger than in any other business segment. Several years ago, I was very proud about the fact that I was appointed to the US Manufacturing Council by the previous secretary of commerce. And I’ve been able to advise member of members of Congress and the administration on manufacturing conducive policy. So to me it’s a meaningful activity that’s contributing something to the society.

VO: Manufacturers can’t avoid hearing that we’re now in the fourth industrial revolution. How did we get here? Steve explains the key stages of manufacturing evolution and what it means to the workforce today.

Steve: The big steps in manufacturing evolution were certainly: total quality management – TQM and subsequently statistical process control. So after that we had computer controlled systems on the factory floor, CNC machines. We saw the connectivity between machines, ERP systems. Another incremental improvements were driven by  certainly value stream assessments and lean initiatives where manufacturers really started to zero in on understanding how value was created on the manufacturing floor and then subsequently eliminating wasteful steps. That was quite a transformational process for many manufacturers. And then today we’re at the advent of Industry 4.0. Which brings the ability to collect data from many different input points, from machines, from processing technology, offline quality operator inputs. Collecting a massive set of data, analyzing that data  and storing that data for future comparisons. And really helping or providing the insight of what is happening on the manufacturing floor. In addition, the insight or the transparency is certainly an enabling feature.

What we have today, we are at the advent of the fourth industrial revolution. And this in my mind is going to be the most transformative, the most significant one. Because we can collect data from many different sources: from machines, operator input, offline quality. And we can analyze that data, make the history of that data available. Then we can provide operators, process technicians, plant management with the inside on how to adjust their processes so they will always operate in the most efficient way. Meaning faster throughput highest yields at the lowest total cost.

An additional benefit of having that inside, having data analytics, is really helping manufacturers overcome the workforce development problem. There is a lack of skilled workers in a manufacturing environment. Baby boomers who have a lot of that domain knowledge and experience are retiring, less skilled, less experienced team members join companies. With the ability to provide process insight we are leveling the playing field to some extent. We are giving the opportunity to less experienced workers to perform just as a high as a level as somebody that might have 25 or 30 years of experience. So this enabling the usage of a subset of data analytics where we really help manufacturers to operate at the highest possible efficiencies .

VO: With background in robotics and automation, Steve has a unique view on the role of robots and how it’s changing in manufacturing.

Steve: I studied robotics – the Holy Grail of automation back then. 30 years ago, it was robotic manufacturing with ‘lot size one’. What that means is that every product that comes down an assembly line is completely different or somewhat different from the previous one, and from the next one. So that’s what ‘lot size one’ means. And you can imagine that this needs a lot of flexibility to actually do that for a robot, with all of the supporting mechanisms to adjust with each cycle to a different assembly or to a different manufacturing task. Automation and robotics prevail in high volume long lifecycle product manufacturing, because tooling up an automated assembly line and that cost is depreciated over the lifetime of a program.

So automated assembly robotics always have been very prevalent and very conducive in high volume manufacturing environments. And what it lacked and still does to some extent is the flexibility to adjust to a different product. So to some extent if you take cell phone manufacturing for example, you have a new model launch every nine to 12 months. Also there is a big mix of different models with access to cheap labor. It typically has been more conducive to have some conveyor assembly lines but still have humans running manual assembly operation. I think the industry is still trying to fulfill again that Holy Grail where you have robotics, where you have automated assembly that is completely flexible and can perform not only a repetitive task but whatever the task requires. So that might be a different shape of components that are being assembled at different locations with different features. So that needs a lot of technological breakthroughs.

For example workers can feel a certain resistance for example or apply just a certain pressure. Instead, robots are vision controlled. So they’re looking for the start and the end points versus having a fixed programmed and started end point including the path between those two points and then also the entire supporting infrastructure. If you assemble a product you most likely have screws, you have inserts, you have circuit boards. They typically all need to be in a jig so that the robot can pick up these components and place them repetitively in the product that is to be assembled.

If you have a lot size one you can not have (…). So these are hard tools automation solutions and the opposite of that is again that you can bring any product, any shape in any location within a certain proximity that the robot can find. The robot understands the location and the orientation of the part that is being picked up and subsequently assembled. So it requires a lot of computing power, requires vision sensors and again –  a completely different concept from a hard coded programmed path and task that the robot fulfills.

A lot of progress has been made already. Vision sensors have become much more cost effective. The resolution of vision sensors has become much better. But thirty years later, there are still very few examples where automated assemblies with a lot size one has been successfully implemented. I mean in a commercial environment. In lab environments – yes, but not necessarily in your typical manufacturing environment.

The innovation cycle certainly is accelerating to some extent. I mean that also follows Moore’s Law, continuing to have access to increased computing power at at higher speeds. But you know, we have been focusing on tasks again  – a robotic assembly, where the other significant part of manufacturing are processing industries. Where were you actually make a part, unlike just putting them together in an assembly environment.

So here you have plastics processing. You have CNC machining, you have dye casting, a lot of other processes that require a certain expertise in fine tuning. So that you can operate at the highest yields, with the highest throughput and the most cost effective way. And that is an area where we have seen much faster progress then with some of the assembly tasks and robotics use that I just described before.

VO: But will the robots take over the factory floor?

Steve: I think that is foreseeable. I don’t have a crystal ball or make any prediction when that happens. But there will always be some level of it. Well, first of all that also really depends on the type of product that you are making or on the type of industry. But you always will need some level of manual human support. One is obviously maintaining the equipment. And here also with Oden Technologies, we are currently expanding our machine learning, our artificial intelligence capabilities in providing equipment manufacturers and ultimately manufacturers users with predictive maintenance.

So we analyze migration patterns, magnetic fields temperature and we able to conclude the state or the condition of components or of subsystems of a machine. Right now one of the largest disruptive elements in a manufacturing environment is unplanned downtime, unavailability in availability of machines. So if we can monitor the condition of components subsystems and predict with a level of accuracy when a component fails, and advise the maintenance team to change these components proactively. Then this will ultimately lead to the elimination of unplanned downtime due to machine failure. So all of these tasks, all of these day-to-day headaches that we have a manufacturing floor, with the availability to collect, analyze and interpret data will lead to a significant changed environment. And again lead to a much more efficient way on running your manufacturing floor.