This interview with Carl Vause, former CEO of Soft Robotics, was conducted by PreScouter Technical Director Ryan LaRanger, PhD.
“With new technologies like soft robotics and collaborative robotics, you are starting to see robotics penetrate into new markets like the food industry in multiple vertical segments including farming, supply chain, and grocery delivery.”
Where do you see the future of robotics moving in the next 5 to 10 years?
The next major advancement in robotics is going to be overcoming the variation barrier. Robots are very good at welding cars and palletizing boxes, but it is difficult for them to operate in environments which have high variation. In many respects, variation is the enemy of robotics. There have been decades of research that have gone into solving this problem. Some potential and very complex solutions which have been explored include the development of hugely complicated algorithms and machine vision. There has also been a great deal of work done in developing complicated grasping mechanisms.
The need for robots which can work well in a high-variation environment is very real and is receiving very little press right now. This is particularly important for consumer-driven manufacturing strategies. Before we got into the omni channel world, customers would go into a store and pick between a set number of options on the shelf. Manufacturers could do relatively straightforward SKU optimization, figure out what product was the most profitable, and focus on making that exclusively. Companies could then spend a great deal of time optimizing out their manufacturing line to make as few of a given item per batch as possible.
With the advent of online shopping, the consumer has all of the power. If they don’t want the item you have on a shelf, they will leave and find the thing they are looking for. This has led to manufacturing de-leaning, or SKU proliferation. We know some companies are talking about on-demand manufacturing: a single SKU for a single customer, if you will.
Companies are turning to automation to solve this problem. If I am a consumer goods company, I am stuck manufacturing all of these proliferating SKUs for different channels, segments, consumers, and clients. These manufacturing facilities are so-called unstructured or low-structure environments, and they are where things are starting to turn, as traditional automation is ill-equipped to handle these environments. This is where technologies enabling automation which can accommodate variability will be so important.
Traditionally, automotive and electronics have represented 75% of all robot sales. Robots are very good at welding and are very bad at accommodating new shampoo bottle shapes for holiday promotions. We have seen the robotics market dominated by automotive and electronics and what we call “general industry.” In the robotics field, we joke that general industry is anything not automotive.
With new technologies like soft robotics and collaborative robotics, you are starting to see robotics penetrate into new markets like the food industry in multiple vertical segments including farming, supply chain, and grocery delivery. If you look at what Kroger is doing, you can see that they are spending hundreds of millions of dollars on grocery delivery automation.
These advancements are really changing the use case, and that is really where the majority of robotics will be moving forward, bringing automation into industries such as consumer goods, where you have high variation in items like shampoo and bottles of toothpaste. For example, cosmetics is a space where there is an incredible amount of variation; it has very fast product cycles. Logistics robots may be able to deliver goods right to your door, but only if they are able to efficiently handle a high-complexity environment.
Right now the process by which food gets to your table is highly manual, and this is a problem because of labor scarcity. In the food space, there is also an enormous degree of variability. We really need to figure this out because the SKU proliferation genie is not going back into the bottle; the consumer isn’t just going to wake up one day and decide they want to go back to a choice of six.
There is a classic axiom in automation which is that the more heavily roboticized your factory is, the more difficult and expensive it is to retool the factory. How do you feel this concept will be challenged moving forward?
Moving forward, how do I build a factory that I can rapidly and economically retool is a hugely important question.
“What we see with the outlier companies who are quickly adopting new technologies is that the initiative is driven from the top.”
How do you see robotics changing in spaces where, right now, it does not have much exposure; for instance, healthcare and urban robotics?
I think there is a paradox that is interesting. These industries constantly say that they are under pressure. And technology is advancing in such a way that they should be rapidly moving towards adoption. What they are seeing, however, is that these industries are not rapidly adopting robotics; in fact, they are often not even piloting.
I am in a working group; what we have found is that often, the burden of the manufacturing process is on the manufacturing organization. They are often resistant to allocating sufficient funds to make the pilot a success.
So you have these innovation teams, and these innovation teams test out and see technology. But they will not be able to buy or deploy the technology. What we see with the outlier companies who are quickly adopting new technologies is that the initiative is driven from the top. Some groups are kind of hoping that this process will bubble up from the bottom, but this does not work. Instead, it leads to a never-ending cycle of piloting and testing and the great leap to production does not wind up happening.
“This is exactly where we are with robotic automation today. It’s a journey.”
Given this, what do you feel are best practices for incorporating advances in robotics into your manufacturing process?
From what I have seen, the companies which are most successful with this are those which have a senior executive sponsor the automation initiative who reports to either the CTO or the CEO. Some of these successful companies have a chief digitization officer or someone else at the C level. Companies which have made automation part of their strategy from the top, with an accountable senior executive executing on this, are generally outliers that are well ahead of their peers. Unfortunately, when manufacturing engineers try to tackle this challenge, they often don’t have the time and ability to deploy and start up the new technology.
Another reason why we are not seeing so much uptake: This is a journey. I often talk to customers; in the 90s we were rolling out ERP systems. Starting up Oracle or PeopleSoft or similar systems was not a fun experience, but we could not imagine, now, running our systems without it. You look back now and say, my gosh, if we didn’t start up an SAP 10 or 20 years ago, where would we be?
This is exactly where we are with robotic automation today. It’s a journey. If you know it is a journey, you can test your adoption of automation and robotics through setting clear milestones with clear metrics and say, if we hit this milestone on enterprise adoption, then we can move forward.
“A big part of the technological challenge is grasping. I think that there is a big challenge in ease of use for automation. Solutions in this space will facilitate the development of truly collaborative robots and some of the bigger solutions like autonomous mobile robots.”
Do you see any specific technological challenges which are preventing adoption of robotics in some of these general industries, and do you see them being overcome in the next 5 to 10 years?
We deal with the grasping problem; once you make a good grasper, you can go full bore. We are deployed today with customers around the world, and things are moving in the right direction. It is largely about making people aware that this is something they can do. Often, people are just not aware that there is a solution out there.
In the case of autonomous mobile robots, you will see increased adoption because the price of LiDAR is coming down. I think that is very interesting.
A big part of the technological challenge is grasping. I think that there is a big challenge in ease of use for automation. Solutions in this space will facilitate the development of truly collaborative robots and some of the bigger solutions like autonomous mobile robots.
Are there any particular enabling technologies in the collaborative robot space which are particularly exciting right now?
The Universal Robot is kind of the big one which has enabled ease of use. Being able to use it out of the box, where you don’t have to go away to a special school to learn how to program it. It has a touch screen, where you don’t need more than a very basic mechanical background to immediately start using it. With collaborative robots, ease of use, ease of setup, and ease of programming are very important things.
Moving forward, how do you see your company’s technology developing?
Traditionally, we have worked on bringing agility and adaptability to industrial robots. Additionally, we are now doing a lot of work in the collaborative space. These are customers who want ease of use, changeovers, and rapid redeployment of their robots. Our technology facilitates this. The dream is to have a robot which can drive around and grab things. Now, a robot which can only drive around and pick up one item at a time is not very valuable to you, so we are working on making these robots more efficient and capable of picking up a broad range of SKUs. We are starting to marry our technology to mobile robots that have arms on them to make this a reality.
