Robots and AI can 10X productivity. Are they also better for workers?

Photo by Jay Wennington on Unsplash

In the future, most work might be done by robots, but right now, most is done by humans. How do we manage the transition? And, what does a humans + robots economy look like?

In this episode of TechFirst with John Koetsier I chat with Lior Elazary, CEO of inVia Robotics.

What we chat about …

  • What is robotics as a service (RaaS)?
  • How can you ultimately get the best contributions out of what humans can do and what robots can provide?
  • What kinds of productivity gains are you seeing from robots?
  • What size of warehouse works best?
  • Does this change how big warehouses need to be?
  • What are we learning now about the future of automation?
  • How do you see the world of work in 10-20 years?

Here’s the story at Forbes …

Scroll down for full audio, video, and a transcript of our conversation …

Listen: robots and AI can 10X productivity


Watch: are robots a net positive for human workers?

Subscribe to my YouTube channel so you’ll get notified when I go live with future guests, or see the videos later.

Read: robots, AI, and e-commerce shipping

(This transcript has been lightly edited for clarity).

Top quote: A lot of people don’t realize how gruesome it is to work inside the warehouse. A lot of times when you end up working with the robot, you’re becoming more efficient and that actually is a really, really good feeling.  

John Koetsier: In the future, most work might be done by robots. Right now, most is done by humans. How do we manage the transition?

Welcome to TechFirst with John Koetsier. There are a lot of political, social, and economic issues to solve around robots. But there’s also a lot of practical issues. How do they work with humans? How do you start incorporating robots? And how do you maximize both human contributions and robotic abilities?

To dive in, we’re chatting with Lior Elazary, CEO of InVia Robotics.  Welcome, Lior! How are you? 

Lior Elazary: Very good, John, thank you very much for having me on the show. 

John Koetsier: Hey, it is a real pleasure to have you. Talk to us a little bit — what is robotics as a service? You offer that as a company. It sounds cool, everything is as a service, platform as a service. What’s robotics as a service?

Lior Elazary, Founder / CEO at inVia Robotics

Lior Elazary: Yeah, so robotic[s] as a service is a business model that allows us to sell the productivity of the robots instead of the robot itself.

So if you imagine in the past, you used to buy the robot and you were in charge of that productivity. If the robot didn’t work, or needed optimization or anything, you had to deal with it because you paid for that.

What we’re doing instead is selling the productivity of the robot. So if you’re buying the productivity, that means you’re buying the tasks that the robot is doing.

And we’re in charge of making sure the robot is working, operational, optimized as well to make the task more efficient. And that really puts the burden right on us, where it should be, right? We are the robotics company, we have the PhD here with robotics to be able to do that. And we think that’s really the — not only the future, but it’s how we’re going to be able to introduce a lot of robots into the workplace, because a lot of our customers are just intimidated by that. So we take care of that for them. 

John Koetsier: That’s super interesting. I mean, it’s the triumph of the subscription economy, right, in every place. And I guess that’s really interesting, because for you it maximizes the ability to develop the solution as you go along.

For a customer, it means that they don’t have an initial $500,000 … $3 million outlay right off the top, correct? 

Lior Elazary: Correct. And not only that, they’re also buying for the need that they have today. So they don’t have to buy hundreds of robots to figure out what to do with them, you know, if they have so many items that are leaving out the door, they are buying — or we are providing those same robots for them.

And as they grow, we’re able to provide them with more robotics, with more systems, more optimization to be able to handle their needs. 

John Koetsier: So let’s talk about how it actually works. And meanwhile, we’re showing some of your robots in operation here, but we have mixed workforces, right? Robots can’t do everything, they can’t pick everything, they can’t grab everything, they can’t pack everything. They can do some things.

How do you maximize productivity in this sort of in-between state here? 

Lior Elazary: So when we looked at the problem, we looked at it in terms of a couple of things, right, that you have to do inside the warehouse. And there’s actually three things, but one of them is software related.

So the two things is movement, right, you’re moving items around. But the other piece is manipulation, and you have fine manipulation and gross manipulation. Humans and people, in particular, right, are really, really good at fine manipulation. I’m really good with my hands doing very fine, intricate tasks and things like that. I’m also good at walking, but in some sense if I’m walking it’s a waste of my energy, a waste of my time, a waste of my intellect, right, what I’m doing there.

So what we’ve done is we developed robots that are able to, first of all, move around in the warehouse autonomously. And then on top of that, they’re able to do gross manipulation, as you can see here, they’re manipulating boxes, they’re manipulating more things that robots can do these days pretty easily. And that allows us to bring the item, as you can see here, to a person who will then do something with it.

For example, packing.

Packing is a really difficult thing to do, because you’ve got to figure out how to intricate and how to move that in. And it’s not just that it’s difficult, it’s also if you want to try to do any value-added services, for example, we have our customers doing embroidery on the items itself.

So the robots are basically doing all the movement, we’re eliminating the movements completely from the people that they don’t have to do the movement. And now they’re doing more of that value-added tasks, and that makes them a lot more efficient.

So instead of them fulfilling a few orders a day, they can fulfill — sorry, a few orders an hour, let’s say — they can fulfill hundreds of orders now an hour, and do that pretty much with more value-added services.

John Koetsier: That’s very, very interesting. And you’ve done this in a number of different places. How’s that gone? How do people react to that? How do they work in that new environment? Does that feel good to them? Do they feel threatened?

Tell us a little bit about how it feels to work with robots as coworkers, because there’s probably a very small percentage of the globe’s population, the globe’s workers, who have actually had that experience so far.

An Invia robot

Lior Elazary: Yeah. So I think a lot of people don’t realize how gruesome it is to work inside the warehouse.

So, I think what’s happening is that a lot of times when you end up working with the robot, you’re becoming more efficient. And that actually is a really, really good feeling.

Imagine again, if you’re in a warehouse walking miles and miles a day collecting all these orders, and you just never have enough time to collect these orders. Now instead of walking all these miles a day, you’re standing still, the orders are coming to you, you’re fulfilling all these orders at such a massive rate that it makes you feel a lot more productive. It makes you feel a lot more satisfied with your job because you’re able to do that. And we’ve seen that.

So a lot of the employees that have used the robots, often our customers have that need, they have all of a sudden this huge number of volume that they have to supply. And instead of the employees being stressed of how I’m going to do that as quickly as possible, they feel comfortable that now they can do that, because the robots are doing most of the work for them. So, you know, we have that.

Now on top of it, we have other jobs, right, that we’ve created — which we call, like for example, ‘robot wranglers,’ right?

Somebody needs to go in the warehouse and deal with the robots, fix their problems. Robots are really actually not good at all about fixing their problems. So we have people who are doing that — both in the warehouse, so we train some of their employees, and we also do that remotely. We have a robotic operation center that is monitoring all of these warehouses in real time to make sure they’re always on, working a hundred percent of the time. And we have a team of people basically using their talents and distributing it across many warehouses, as opposed to having, you know, hiring [a] PhD to operate inside your warehouse.

John Koetsier: That sounds good, and that sounds interesting. And that’s probably a positive there in terms of the future work life of just about anybody who deals with things that we’re building or manufacturing or shipping, because that’s going to get increasingly roboticized over the next decade or so.

It’s interesting for me, because when I was in university, I did work at a warehouse. In fact, I did the overnight shift — the Friday night shift from like midnight till 7:00 AM, went and played my ice hockey game the next morning, right after, enjoyed that — couldn’t do that these days. But it was a chaotic disaster there, right? There was stuff all over the place, nobody knew where anything was. It was common for somebody to disappear for half an hour to find something.

You offer a warehouse execution system that somebody can purchase even without buying robots, correct? How’s that work? 

Lior Elazary: Yeah, that’s correct. And in fact, you talked about the night shift, we have a few customers that do that and they say this is where all the shenanigans, you know, happen. And it’s— 

John Koetsier: Yes. Very, very true. Very true. I can personally attest to that. 

Lior Elazary: So it’s really because — and it’s actually not the employee’s fault, right? It’s really because the work, it has not been organized. So, you have lots of different work that you gotta do, and it’s very overwhelming for a person to realize, okay, do I do this right now? Do I do this right now? Do I go and pack? Do I go pick this order?

You have SLAs, these are service level agreements that you have to make sure those orders go out the door at this time. But you gotta make sure is these orders, is FedEx coming? So there’s all this work that has to be planned.

And a warehouse execution system is basically the conductor of the warehouse. And we’re offering, in fact, just that software, and we’re seeing almost 50% increase in productivity just by introducing that. So you as an employee, you come in, you log into your terminal which is on your wrist, and automatically it’s putting you to work, and the system coordinates between everything. So if you go to a location and, for example, it’s short — the item is not there and you’ve got to short that order — it will automatically dispatch somebody else to go to a different location and get that so that order can get out the door.

And that is actually a very, very complicated mathematical problem to solve. But we have a computer system that actually we do that, run that on a cloud, that solves that [and] takes into account not only what you’re doing today, but probabilistically try to determine what’s going to happen in the next few steps in the next days, and set up some stuff  for that. So that actually is something that really provides you with a lot of efficiency very quickly. 

John Koetsier: Interesting. So I can imagine that system taps into what it knows are orders that are going to be going out, what it knows about supplies that are going to be coming in, and then orients those, and positions them in the warehouse where it makes the most sense.

You mentioned a little bit briefly about it, but maybe you can go into more depth. What kind of productivity gains are you seeing from robots, maybe from having the system in the first place, but also having the system and then the robots, and as well having the AI applied? 

Lior Elazary: Sure. So we often see about — just by introducing the system itself — we usually see about 2X productivity. There’s a lot of, as you know, a lot of idle time, right, that when you see when you worked at a warehouse, a lot of time that, you know, just moving around not actually doing the tasks they’re supposed to do.

So just with that, we’re able to see 2X.

Now introducing the robots, now completely eliminates the walking and we’re able to see 5X, and in fact, with some customers we’re able to see up to 10X in productivity gains. Because now, instead of the person having to walk around collecting all these items, the items come to him and he’s able to operate a lot more efficiently.

Now the AI — and a lot of people hear that word and are like, ‘Oh, what does it do? What does it do? But it’s really there to adapt to changes. If things in the warehouse were constant and never changed, we can develop one piece of software to execute and to do these things over and over again. But the reason we need AI is because constantly in the warehouse there’s a lot of changes — the truck didn’t arrive in time, you have a bunch of orders. There’s a predictability portion to it, you know, is Elsa gonna be the next thing that’s gonna run out in the next few days? That’s what happened last Christmas.

So those things [are] what the system, the AI system is capable of doing. And then it’s able to replan and recoordinate the tasks around in real time to get those orders out the door and to be as efficient as possible.

John Koetsier: Yeah, that makes sense. What kind of costs are we looking at here for, let’s say, an average size warehouse? What does it take to get there? 

Lior Elazary: So, because we have the software component, we can actually start with a very small warehouse. We have systems that start at $2-3k a month, and then they just grow from there, right? So we can grow with the customer, grow with his needs. Grow as they’re producing more and more, you know, their volume increases, we’re able to do that.

And in fact now with COVID, we had several customers that saw larger peaks of this year than they’ve seen of last year during their actual peak. And if you asked them back in 2019, that in May they’re going to see huge volumes, they would’ve think you’re, you know, smoking — like it would have been, pretty wild.

But that happened. So what we’re able to do is actually being able to support them and provide them both with more robots. We actually enhanced some of the algorithms, some of the efficiency to give them more productivity. And they were able to sustain and actually provide all of that work that they needed to do all that volume, all the things that people ordered, with a shortage of labor, right?

So the other piece is that labor just wasn’t there.

John Koetsier: This is really, really interesting. And everybody knows that our economy is changing as a result of COVID, and that’s obvious. And we’ve seen four to six years, that was the estimate from, I believe, Salesforce — or actually Adobe. We’ve seen four to six years growth in e-commerce just in a couple of months during the early phase of COVID here.

But that’s something that I hadn’t considered as well, is this extra productivity is wonderful. That’s great. But you know, you couldn’t get labor in, you had other challenges around COVID, you needed to separate people, right? You can’t have them working in close proximity. So this is actually an opportunity as our economy changes and shifts to a more e-commerce ship-it type of economy, this is perfectly situated. 

Lior Elazary: That’s right. And, you know, we had a lot of retail stores, right, that closed down. And that really just put all that pressure on these warehouses that really weren’t geared for it. You know, even Amazon sort of slipped, right, there was no more two days, it took several weeks right in the beginning.

So it just, it shows you how complex that problem is. And if we really wanted to move toward like an e-comm, basically solve the logistics in a much more optimized manner, as opposed to us just walking around the stores and us being the pickers for ourselves. It’s still, you know, something that we need to solve.

And the way we’re going to do it is, we’re not going to be in a position where half the people in the world are going to be ordering and the other half are going to be working in warehouses being our personal shoppers.

So the way we’re going to do this is by making a person a lot more productive so they can fulfill many, many orders. And the way to do it is with robotics, right? And…

John Koetsier: This is exactly like the productivity gains that we saw maybe in the turn of the century, the previous century in farming, right?

Lior Elazary: Yep.

John Koetsier: Half the world was a farmer. 

Lior Elazary: Yep. Exactly.

John Koetsier: Half the world lived on farms, and they made food for the other half and themselves. But, then with tractors and other machinery, they could actually, you know, now it’s what, 3-5% of people—

Lior Elazary: Yep. 

John Koetsier: Live on a farm and do a farm. I want to ask, does this make warehouses more effective per square foot as well? Can you have a smaller warehouse as a result of putting a system like this in place? 

Lior Elazary: Yeah, definitely. We’ve actually created this system to be very highly optimized towards smaller footprints. And the reason for it is that, you know, if you’re ordering something from New York and you’re in California, you’re not going to get it within an hour. So the only way to get it within an hour is to have warehouses that are located everywhere. But they’re not going to be a million square foot.

You’re not going to put a million square foot warehouse inside the city. So you gotta be very optimized.

And that actually goes in, again, with the warehouse execution system where we’re able to both optimize the space that you need, moving, defragging things in real time, the robots actually do that overnight, and optimizing the space as well as the throughput. 

John Koetsier: That is really, really interesting. Does the execution system — is it multi location capable so that, for instance, if you had a large distribution center for a region and then you had small ones in cities or other areas like that, can it handle traffic between those?

Lior Elazary: Yeah, we do. And in fact, we have 3PL third-party logistics who are our customers who have different systems, and [they] have different warehouses across and they use our system for the, basically their warehouse execution. So we can decide, you know, which warehouse should we deliver this? Should we move items between warehouses? And so forth.

So it really it’s more of a global optimization in how do you get the best delivery times to the people, right? Because in the end, we’re an on-demand society, right, you want your order, you click add-to-cart, you want it right now. So you need that infrastructure and that warehouse execution system is deciding how to best achieve that. 

John Koetsier: And the whole time your AI system is seeing what’s happening, optimizing what’s happening, getting smarter and smarter over every time it makes a shipment, or adjusts inventory or anything like that. 

Lior Elazary: Exactly. So adapting to changes, right? So all of a sudden a truck didn’t show up because it was an accident on the freeway, right, and you can’t, you didn’t get those items in time.

So you might dispatch it from a different warehouse that’s nearby, and maybe it’ll cost you more, but you’ll still be able to get it in time. So these kinds of real time decisions, you know, that’s where the AI is best suited for being able to adapt to these kinds of changes that you just can’t predict them ahead of time.

John Koetsier: What are you learning as you’re running this company and dealing with customers, about the future of automation and commerce and shipping. You’re doing a lot of shipping and receiving right now. You probably see that integrating with production in the future.

What are you learning about what our future’s going to look like?

Lior Elazary: So, I mean, I love it. So right, as we’re seeing from our customers, we’re striving to [be] able to provide items, right, to people’s hands as quickly as possible. And there’s a couple of things.

One is the multitude of products.

So I know a lot of people told me, well, you know what? I still want to go to the supermarket because I want to pick my own avocado. But if you have somebody who picked always the best avocado for you, you would just order it online, because going to the supermarket wastes about two hours out of your time. You gotta go collect the items, go back to your house, put them in together.

So as the future progresses, and we’re able to move these items and get you the items that you want at the tip of your fingertips, it both as a society makes us extremely efficient, but it also opens the door to a lot more things that we can do. For example, our company, we’re able to iterate the robot very quickly, because we’re able to order parts online and get them within a day. If we have a motor that we’re testing and it took, you know, previously it would take a month, right, to get that motor in. It will be a month out of the time that we couldn’t iterate the robot.

Being able to get that motor over a day, we can iterate it, build it, see okay, that didn’t work, let’s get a different one. And then in a couple of days we find a perfect solution. So it’s really just, it’s not just about getting the toys, it’s also about getting a lot of stuff and being more productive as a society. 

John Koetsier: Last question. There’s always a lot of angst — and there was when I shared this to Facebook as a broadcast that would be coming up — there’s always a lot of angst about the future of work and where people will be.

Will they have a job? What will that look like? I want you to kind of take out your crystal ball, prognosticate a little bit as somebody leading this industry to bring more robots into the workplace. How do you see the world of work in 10 to 20 years? 

Lior Elazary: So, I think, again, as a society we’re just going to become more and more efficient. And with more efficiency comes more responsibility, right? So we can make our choices of what we want to do, but ultimately we’re going to create a lot more jobs that we have.

If you look at what happened 15 years ago, right, being a search site optimization person, nobody would’ve ever thought that that is a job that actually exists. And same thing is going to be here. We’re going to be able to create so many more jobs, and people are going to be able to do so much more.

Again, we can draw parallels from the music industry, right? We completely solved the logistics problem of distributing music, right? I have music on my phone, whatever I want instantaneously I get that. I don’t have to wait for the CD to come out or the record player to come out. But, what it also did, is it enabled people to be more creative. So we have a lot more artists creating music. We have a lot more — we’re consuming a lot more different types of music. We’re exposed to different things.

I think the same thing is going to happen with items. So as we’re able to give items, basically access to items, and solve that logistics problem, basically get as close to the Star Trek replicator as we can, right? I want this item, it’s available at the tip of my hands.

That will be really powerful because now you’re going to have more people creating more things. People are going to have better jobs, right, as we saw through technology over the many years. So instead of like you mentioned with the farmers, so everybody being farmers, we have bankers, we have a service — I think we’re going to move a lot more towards a service-oriented type of industry where we’re really using people for what they do best, right?

People were not built to move around warehouses and walk day in and day out. We’re really meant to think. We’re really meant to create, to enjoy our life.

And I’m looking forward to it, ’cause I think that’s what it’s gonna do to us, right. It’s going to give me a lot more time in my life to do other things, creating companies. And you know, I’m really excited by that. So I think it will be a great future. 

John Koetsier: That’s a good answer. I want to say ‘Tea, Hot, Earl Grey’ —  but you’ll get the reference since you mentioned Star Trek

Lior Elazary: Yeah, exactly. 

John Koetsier: Exactly. That is interesting, and I hope that the future does unfold something like that, that new niches do appear as we solve problems. And if you look at the data on sick days and injuries for Amazon workers in their warehouses, that makes some sense as well. Lior, I want to thank you for joining us on TechFirst. It’s been a pleasure. 

Lior Elazary: Thank you very much, John, I loved being here. 

John Koetsier: Excellent. For everybody else, thank you for joining us on TechFirst as well. My name is John Koetsier. I appreciate you being here. You’ll be able to get a full transcript of this podcast in about a week at The story at Forbes will come out right after that. Full video is available on my YouTube channel within minutes, and actually the audio podcast is available almost immediately as well.

Thanks again for joining. Until next time … this is John Koetsier with TechFirst.

Subscribe for free

Made it all the way down here? Wow 🙂

The TechFirst with John Koetsier podcast is about tech that is changing the world, and innovators who are shaping the future. Guests include former Apple CEO John Scully. The head of Facebook gaming. Amazon’s head of robotics. GitHub’s CTO. Scientists inventing smart contact lenses. Startup entrepreneurs. Google executives. Former Microsoft CTO Nathan Myhrvold. And much, much more.

I’d appreciate it if you’d subscribe on your podcast platform of choice: