Hi everyone and welcome to another episode of Growth Challenges for Manufacturers and How to Overcome Them. I’m your host, Katherine Seymour. This episode is brought to you by MAcRAE’S, trusted by North American businesses for over 100 years. As a leader in B2B digital marketing, we help industrial and manufacturing companies struggling with stagnant growth or lead generation. By leveraging advanced AI integration and automation into our SEO and lead generation programs,
We help you appear more prominently on search, AI results, like chat GPT. This drives significantly more traffic to your website, resulting in stronger lead flow and increased revenue. Crucial in today’s highly competitive digital landscape. Learn more at macraes.com today. Today I’m joined by Ville Lehtonen VP of product at Real-Time Robotics, a company transforming the way robots plan, move, and work in dynamic environments.
Real-time technology enables industrial robots to generate collision-free motion plans in real time, allowing manufacturers to achieve faster deployments, greater flexibility, and a safer collaboration between humans and machines. By combining advanced motion planning and optimization tools, they’re helping manufacturers reduce downtime, increase productivity, and unlock new possibilities in automation. It feels so great to have you on the show. Welcome.
Thanks for having me.
Thank you so much. Let’s get started with your background. Can you tell us a little bit about your career journey and what brought you to real time robotics?
Yeah, so I am a computer scientist to begin with and ⁓ initially I grew up in Finland so I graduated while Nokia was still doing well so I did telecoms for a little bit. ⁓ But then I went to study at Oxford and from there I founded a startup and doing robots for life sciences. So I’ve been doing robotics now 17 years which is dating me a little badly here.
Yeah.
Long time.
First, Robotics and Life Sciences. Then, migrated, sort of exited that company. Then joined a friend’s company during COVID, which was also doing life science. were doing all the, setting up all the testing for California, but also Moderna developed their mRNA vaccine on the system we delivered, which was very cool. And then I left that and joined a company called Pickle Robot to run their product management to…
⁓
Mm-hmm.
sort of truck and they do truck and loading which was very very interesting to do but then RTR came not calling. RTR is very well known in Boston for having some amazing technology and I thought it was a very interesting challenge because
Okay.
The technology is incredibly good, yet it didn’t have an enormous commercial footprint. I took it as a very interesting challenge to kind of go see, it’s like, well, what all can you do with this technology? And is it really quite as good as everyone says? And the short answer to that is yes, it’s some of the best tech I’ve ever seen.
Mm.
Amazing. As a VP of product, you’re like at the intersection of technology and customer needs. How do you approach shaping the product roadmap in such a rapidly evolving field?
So the interesting part, this depends a lot on the company. Sometimes, like it was very fascinating. Pickle, for example, was 100 % customer centric. Customers need to unload trucks and the company’s kind of living purpose was like, how do we unload a truck, right? So it came from one direction to the other. RTR is very different. Real time started from an incredible innovation where they kind of, there were two academics at Duke.
I ain’t.
Yeah, yeah.
And there was this common problem that everyone was trying to solve, like Siemens and Kuka and these big companies. And they came up with a way and they’re like, how much faster are we going to do this? This takes 10 seconds normally. we took two milliseconds. we were over thousand times faster than the state of the art. That’s kind of cool. And it is cool, but now it’s going the exact opposite way. Like, people need to unload trucks. It’s like, we did a thing in two milliseconds, which is…
Wow.
Right.
Interesting, but no one’s going to buy a thing because of that. So where my role gets very interesting is that it’s very different in those two companies, right? And the other one is making sure that the tech team is focused on the right thing. So you’re kind of internal facing and making sure it’s like, don’t solve problems that the customers don’t care about.
For sure.
Mm-hmm.
Right.
That was the Pickle experience. At real time, the experience is more like, okay, this is super cool and this is super powerful, but whose problem does this solve? And one mistake that people very easily do is like, well, this solves the problem that Ford is having. Who’s Ford? That’s not a real thing. That’s a company.
Right.
It doesn’t help if you solve a company’s problem. The question is who inside Ford’s problem it is. Because if it’s 1 %… I learned this the hardware at my startup, which we solved. But if you are really nice, there’s 100,000 people who want your product for their Christmas present and it’s number 7 on their wish list. You might not sell the single product.
Right.
Right, right.
It
would be a lot better to be number one for 50 people than it is to be number seven for 100,000. And that sort of boiling the ocean problem is like you need to find the exact person for whom you can be number one. And that’s the sort of RTR challenge has been is like, okay, motion planning. Who has a problem with motion planning? I don’t have a problem with motion planning.
Right.
You run a factory, do you a problem with motion planning? That, you have to position it correctly where it resonates with the buying authority. It’s like, oh, that’s what it solves. Okay, that I do have a problem with.
Makes sense, makes sense. For our listeners who might be unfamiliar about real-time robotics does, can you describe what the company does and what it’s about and how it helps the industry?
So this is a cool example of the product management problem in a nutshell. I can tell exactly what technology it does and everyone’s like, well, I don’t Cool. So what it does is you have these complicated robots. There are very many ways that the robots can do what they need to do without colliding. So the number of paths from going here to here with a six degree of arm of fear is borderline infinite. And we deal with work cells that have like 12 robots.
Yeah
So what do we do that’s so valuable? It’s like, well, we just solve it like that.
Okay, it still doesn’t sound that impressive. Well, if you go to someone like Ford and say, hey, we need to solve this 12 robot cell, it takes almost 70 hours of an engineer per robot to solve that cell. So we’re like, okay, so that’s 840 hours and we take 30 minutes.
Bye.
And immediately you’re like, well that saves a lot of money probably because you don’t need to use an engine. That’s the smallest value here actually. But the real value is, and for the car manufacturers, is that the Chinese are coming up with a new car every two years. Westerners, and I’m using that term very broadly, Japan and Korea are on board, it takes almost five.
Bye.
The lead time is the real problem. And also there’s this massive problem that, okay, I did the 840 hours and my robot cell works. Fantastic news. I have 50 of them. And someone comes back and says, ooh, during testing it turns out that everyone that tested this car would like this change.
It’s kind of a big change. have to change… Do you have any idea how big that ask is? I have to update every robot cell. And it all takes… You’re asking me to delay the launch of this car by three months. Are you insane? We’re not doing it. With real time, I’m like, okay. Do the change.
Right, right.
Over next night, we will solve all of that for you and you can… It turns hardware to behave more like software. It’s not the sort of like, you change something, that will take four months. You change something, it’ll take overnight to solve. And that’s the real, real value that we provide is that… And this allows much smaller companies to use robots. Because where the big companies are using our speed…
Right.
to run a lot more motion planning, what we really do is we reduce the cost of programming your robots by two orders of magnitude. So if it cost you 10,000 before, we can do it for 100. You can now do it, you can choose to do it 100 times, but if you’re a smaller company,
Makes sense.
Now suddenly that formidable robotics project that you went and asked for and you said, oh, the robot, the cobot is $20,000. This is going to be awesome. And then you go and you go to an integrator and they are not ripping you off or anything. This is the real cost. And they say, we’ll install it and we’ll program it to do what you want for $250,000.
Not buying a-
Right, right. So really you’re helping not only really big companies who need the speed and the accuracy, but you’re helping the smaller, maybe small, medium-sized companies who maybe didn’t have the budget for it, but it would be a really good ask. They’re suddenly players in this field too, which is really great.
That’s the big deal because there’s this high mix, low volume thing. So robots are really great if you never have to change what they do. But this is exactly the problem because if you want to change the car model you do, you have to take two months every time just to robot program. And there’s a big robot manufacturer who did this study where they kind of discovered that the small and medium companies can pay maybe three, four, maybe $5,000 a year to reprogram their robots.
but they need it done 50 times. You’re paying for the Uber ride to your factory for the robot programmer with that $100 bucks per reprogramming. That’s not happening. With us, it could. It’s fine. We can do it. And that’s a big game changer. We haven’t really done the marketing push for that because we’re kind of working on the higher end of the market. But like to me, the really exciting part about this
Bye.
True. True.
Right.
is this massive democratization of robots because of the cost of that repair. Doing a hundred repair, mean, it’s probably the real cost is 50, 60, 70,000 a year. Who cares if you got the robot for 20,000 at that point? If you get the robot for free, like that was the robot manufacturer we spoke with. Their ironic comment was that they kind of ran the numbers. If they gave the robot for free and gave some of their customers 50,000 as a sort of dowry,
They wouldn’t take the robots anyway.
Yeah, yeah, makes sense.
They’re like, it’s kind
of frustrating because it’s out of our hands. If we make the robots cheaper, everything we have says it makes no difference. None whatsoever because we’re not the real cost of them anymore.
Makes sense. sense. Robotics and automation is a quickly growing ⁓ industry, but adoption still can be challenging. What do you see as the biggest barriers for manufacturers facing implementing robots? Is it really what you just mentioned in regards to the programming and the price it would be? Yeah.
Essentially.
Because the people who have like, well, I’m running in three shifts and I change what I do once a year, they all have robots already. It’s basically that’s the defining feature. It’s like, you go like, oh, we have two shifts and we change twice a year. Okay, well, probably I still have. Okay, one shift and we change three times a year, it starts being people stop having it. And the smaller the company, the less likely you might say, well, I reprogram 10 times and I run…
Great.
factory floor runs 60 hours a week. It doesn’t make sense. The economics just stop making sense for you quite rapidly over there. And there’s 260,000 factories in the United States. 93.1 % I just happened to read this very recently. I don’t actually remember that. are fewer than 100 people in them. The odds that they’re doing double shift never changes products are very, very low.
Mm-hmm. Yeah.
So there’s this whole 250,000 of those factories probably have no robots.
Right.
And
that’s, but it’s purely that people are economically rational. Like if you do three robots and you run 40 hours, we did a little calculator for this. It replaces, it makes sense, but you’re paying $57 an hour for that robot for
So if you can hire someone for, I don’t know, 60 even, it probably makes sense because after all the human is very versatile. you know, whereas the robot runs into trouble if you need to relocate, there’s lots of ways the robot can, you know, cost the building burns or something. The human is far more versatile, so 60 bucks an hour is a lot of money.
Right.
Stuff where it hits $15 is somewhere in the zone of 80 hours a week and 12 robots. That’s a big leap. And a lot of people are not willing to take that leap because the capex for making that leap would be in the $5 million range and you’re not sure if it’s gonna…
Right. Right it is.
Makes sense. Has the conversation with manufacturers shifted in recent years, especially with labor shortages, supply chain pressures, or the push for more flexible automations?
It’s shifted, but for none of those reasons really. It’s because we talk a lot with a lot of robots that are car manufacturing and they are looking for podcast friendly way to say this. They are troubled by the Chinese developments. And they are very, very troubled indeed. Because you look and you’re like, oh, I wonder if I want to buy a 200 mile range EV.
Hahaha
Right.
29,000 Nissan Leaf is the best West can offer. BYD Seagull, similar stats, $8,000. I wish us good luck in the Indian and African markets trying to sell the three and a half times more expensive version of the same thing. This has caused a lot of consternation. It’s caused a lot of, which is, from our perspective, little cynically, it’s kind of good because a lot of companies are quite self-satisfied. It’s works, it’s working.
Yeah.
Right now they’re like, no this is not working. This is really not working. We need to move a lot faster and we need to figure out how to move a lot faster.
E-commerce has changed customer experiences and expectations for the B2C space. Are you seeing similar expectations or speed and adaptability changing in the perspective for your clients?
Kind of. mean there’s a lot of this, it’s not really… Well it is customer facing again down the road. Like the $8,000 versus $29,000. You need flexibility because the world is changing quite a bit right now. So everyone needs to react a little faster. There is also on the logistics side, you can smell this sort of labor shortage issue.
Right.
Yeah.
There’s a lot of people come to us about palletizing in particular. A particularly mixed case. If it’s single case where you’re just packing the pallet, again, robots are pretty good if you’re like, well, I have this one box and I need to bake this pattern on this pallet. Solve problem, most people have robots for that. But most palletizing is not like that. It’s the boxes are a little dented.
hundreds of camera companies around these days to kind of help with this problem. But you can feel all these pressures and all these integrators who build this, they’re getting pretty tired of coding because they are hoping that, and which benefits us, is that there’s thing that, if you build the robot system, you have two things usually. You have the sensor suite of like, well, what’s going on? And the second one is how do I tell the robot what’s going on? And the progress on the sensor suite side in the software libraries has been absolutely incredible over the past five years.
Like, five years ago it was like, oh I wonder if this vision thing can recognize a box from a ball. Now it’s like recognizing a tennis ball from a basketball from a soccer ball and it’s not even having a bad time. Even if you get one of those really crazy World Cup soccer balls which don’t look like a soccer ball at all, it still recognizes. This stuff has come such a long way. The problem is the kinematics has not really progressed much at all.
Bye.
Ha ha!
along the side and this frustrates some people because the vision of PIPELCOM is like, hey, it recognizes everything and they’re like, sweet, I’m gonna solve all the things. But then they get stuck because the robots are different, there’s a lot of complexity on the kinematic side and that’s kind of what we’re hoping to solve is that you don’t have to worry about any of that, you just tell us what you want the robot to do and we’ll guarantee it won’t collide and we’ll guarantee that it’ll move there in a very fast way.
Right.
Very cool. Real time is known for real time motion planning and collision avoidance. How does this technology change the game for manufacturers compared to traditional robot planning?
For manufacturers, we actually have two versions of the product. Now, one we lifted into the cloud, which is not so much real time. This is for factory. This is the one that really reduces cost. What it does is you can say, just like, ask the cloud, how should this work? That works mainly, that doesn’t work if you’re fully dynamic. So if you’re say, palletizing and you don’t know what’s coming down, okay, you can’t run that in the cloud.
Hmm.
Right.
But most manufacturing is either static or what I’d call semi-static. So you change what you do twice a day, even. It’s pretty dynamic. But you can twice a day, if it takes three minutes to get the response from the cloud, that’s fine. If it’s three minutes between every box you put on the ballot, that’s not fine. So this…
Right.
Yeah.
Huggings changed the game largely by that price drop that I mentioned earlier. But it also, because it allows you, the economics start working for a lot of things that no one really thought to do with robots. It’s one of those things, like it’s kind of like, what would it mean if I told you that electricity was, you know, a megawatt was 18 cents now? You’d be like, well, my electric, first thing is like, wow, my electric bill would be a lot smaller.
Great.
Then I’m like, wait, actually, I think I could heat my yard at that point through the winter. The more you think about it, the more you’re like, wait, if the price has dropped that dramatically, that changes everything. And that’s what it ultimately comes down to. Everything, like I said, we could do two things. Because we drop the price.
Right.
to one hundredth of what it used to be. Some companies, like the auto manufacturers who are in a hurry, will do a hundred times more path planning and pay the same amount again. But the others, just, they’re like, well, I’ll take my 99 cents and go home because I’m not in a huge rush because I’m being, you know, under attack from stiff competition.
Yeah, yeah.
that makes sense. Can you share an example of how your technology helped a customer optimize their output or reduce a deployment time?
Um, yes. There’s a couple of ways. So there’s some on our website and stuff where you just have different levels of problems. You run the work cell and you have had people working on it so long, but it is the slowest work cell. So often these are in a row, right? So you say, well, if there’s 10 of these and first one takes 50, 50, 50, 50, 50, and then there’s a 70 and then there’s another 50, you’re like, well, now the whole thing is 70.
right, because they all have to wait. So we have done, can’t name the customer, but it was a Japanese company that you would absolutely know by name. We took one of their sort of sticker cells and after six hours we had taken something that took 102 seconds to 75.
Thank
Wow, that’s really good.
We’ve also worked with others who this turns into a money saving quite quickly. When we took another one at a large, similarly recognizable company just now and we helped them move the robots around. This is kind of insane because a lot of the past just changed completely. So if it takes 500 hours to plan this cell, if I try five different robot setups, I might take 2,500 hours. So everyone’s just like, we’re doing it.
with these robots where they are. And just by moving the robots, because it was pretty obvious when you looked at the cell they were designed, is that there was a bunch of robots and the middle robots were busy like 80 seconds. The front and back ones were both busy in the 25 second range. like, you should move these a little closer because it’s really ridiculous that you have these two workhorses and then you have the slackers on the side.
RING
you
Yeah.
But again, no one wants to commit a month of lead time and 500 hours to it. With us, they’re like, well, let’s just try it out. And we dropped it by almost a third. And in another example where they were like, well, we misunderstood this. We’re actually beating the bejesus out of our target already. We said, well, take one of the robots out. This also happened last week, I think. They took one of the robots out, moved the others a little closer to each other, and it still hit the timer.
⁓
Yeah.
That’s six figures of savings right there, using a couple of hours of cloud time.
Appreciate it.
for sure. Looking ahead, what trends in robotics and manufacturing do you see in the next five to ten years?
I think the big trend, I mean manufacturing, I think there’s going to be this continued approach to trying to make software, hardware behave more like software.
Bye.
The digital
twins omniverse is all a huge part of this trend where the idea is that what if continuous integration, if I code, I’m a coder, I change something, it just kind of is absorbed into the product or I might get a ding back and say, well, you failed all the tests, you suck, try again. But that’s not at all how hardware works.
It’s much harder and there’s this whole cycle and it takes months very easily where they validate and then they push it in. If you have a true digital twin and you have something like our system, there’s nothing to really stop that if someone goes and it goes into the car and adjusts the car.
We can’t do this ourselves, but we are a very important part of this. It would be possible to say next night, do what software does today and say, okay, well, work cell 10 got kind of difficult with that change, but let’s relocate stuff from it. And then next morning you say, okay, the car is now one second slower to make, but we absorb the changes.
I need
and you could do that, that’s going to be a huge change. And I think the other big trend is going to be is that we know we need to sell more robots, right? 500,000 robots sounds like a lot until you put it into context of like just the aging out workforce. Korea is going to lose 400,000 working age people in a tip average year for the next 50 years. China alone.
Yeah.
I was looking somewhere in the late 2030s, and they know this already because if you take the people who turn 66 versus the people who turn 19, they’re going to be dropping the working age population by somewhere in the range of 13 to 14 million a year. 500,000 robots is nothing. It should be 5 million. It should maybe be 15 million.
brain.
Mm-hmm.
compared to that for sure.
There’s going to be a lot of drive to enable that because otherwise we’re going to have very serious problems. US is a little blessed on this. We’re not in as much trouble as everyone else is. But the others are going to need robots so very, very badly to keep their industries going. And the current cost structure is just not okay for that.
Mm-hmm.
Mm-hmm.
What advice would you give to manufacturers who are exploring automation but are unsure how to start or scale effectively?
is that there’s this pessimism slash fatalism. Some people don’t approach it because they hear, like I said, buy a 20,000-dollar robot, it will cost 250,000 to do the software. Some people probably heard that and said, a little rich for me, I’m not even going to bother. Okay, don’t be that pessimistic. It is true, but a lot of technologies, we’re hardly the only ones, but a lot of the integrators are…
No thank you.
slower, so go to the integrators, have a number in your mind, calculate where your return on investment point is before you go in there, and they will say nice things, but say, hey, I would like this, and this is how much money I can, where it makes sense for me. They will improvise. There’s a lot of smart people out there, and if they say, I can’t do this, well, you didn’t waste too much time.
right.
But don’t like go just go and try it out and ask them what are the newest things you’ve seen coming? Are there things that will reduce cost? And this is where the integrators you have to be aware they’re consultants ultimately. And again, I’ve done consulting myself nothing wrong with it. But you never are if someone’s like how many hours do you need to do this? And you know, I mean, I don’t mind saying a larger number. Right. So just tell them very clean cut.
Yeah.
what you need, what you can afford, and you might be surprised at what you get.
Amazing. Thank you so much, Ville. Thank you so much for joining us today and sharing your perspective and how real-time robotics is helping manufacturers today. Your insights on technology, adaption, and challenges has been incredibly valuable. And thanks for everybody being here and listening to the episode of Growth Challenges for Manufacturers. Thank you so much.
Thanks, was great being here.
Contact us to explore how we can help your business grow.
