Robotics can be a controversial topic as the advancement in automation is seen as a threat to employment. “Will robots take our jobs?” is one of the most common questions people ask as robotics and automation begins to increase ROI for businesses.
Beth Kindig interviews Kass Dawson from Softbank Robotics and asks this question and more in Tech Lightning Round’s newest episode. Kass shows Beth two retail robots that co-operate by exchanging IoT data in the cloud named Pepper and Tally, which as Kass points out, removes the human from the equation. Hear how Kass answers to whether robots will take jobs by listening to the podcast on iTunes or Spotify – links are below.
The podcast episode also explores some of the more critical roles that robotics plays, as well. An interview with John Deere reveals that as the population grows towards 10 billion people, with nearly 70% preferring urban locations, growing food will become a substantial problem. Zack Bonefas, a Systems Engineer at John Deere, also discusses how the company plans to use automation to feed the growing population.
Education is also another crucial field where progress is being made with the help of robotics, especially with children who may have learning disabilities. Richard Margolin of Robokind explains how autistic children engage with robots 40 times more than they engage with a human teacher. Richard also manages Robots for STEM, which has helped more children become interested in computing programming at a young age.
00:29 Beth Kindig: Welcome to Tech Lighting Rounds. I’m your host, Beth Kindig. This podcast interviews key people with deep expertise on one topic for a 360-degree view. One difference between this podcast and the other podcasts you listen to, is that I hold short interviews called, Lightning Rounds, with the goal of giving you a lot of compelling information very quickly so you can get on with your day.
00:57 BK: In this episode, I went to CES and met with leading experts in the field of robotics. This topic can be controversial, as on one hand, we want technology to improve our lives and make our jobs easier, but on the other hand we fear the unknown, of what will happen once machines begin to think and act like humans. In these lightning rounds, I speak with experts from SoftBank Robotics, an arm of the Japanese conglomerate, with deep pockets who is taking on retail.
- 01:23 Jill Sciarappo: What level of best practices or validation do we need to have these companies prove they can accomplish before we let them on our city streets?
- 01:37 Michael Fleming: Some of the reasons that mining and defense were early adopters of this technology was, well, on the defense side when we were at war about a third of our war fighters were dying from IEDs or roadside bombs.
- 01:48: Anuja Sonalker: And cyber security is an art, also it’s a technology, it’s a capability. Cyber security is something that you have to be groomed for… It is a mindset. You have to think like an attacker and you have to think like a defender.
01:26 Kass Dawson: The intent is not to steal jobs, it’s to replace tasks. And a lot of the tasks that are being automated are those menial tasks that most employees don’t wanna do anyway. I think it’s a quarter of time of a retail employee is spent on tasks that they just hate doing.
01:42 BK: I also speak with an altruistic company named RoboKind, who is helping to improve the learning abilities of kids with autism.
01:49 Richard Margolin: The robots are really more effective, because they’re predictable. The things we do are things robots are good at, in terms of the way we teach, so it’s perfect repetition, consistency, never getting tired.
02:03 BK: And I also speak with John Deere who explains, perhaps the most pragmatic reason robotics are necessary and that’s to grow our food.
02:11 Zack Boniface: By the year 2050, there’s going to be 10 billion people on earth, and we’re gonna need to grow about 50% more food than we’re growing right now, but the number of farm-able acres that are out there, that’s not increasing, in fact that’s going down.
02:27 BK: My first guest is Pepper, a 4-foot humanoid robot that is built for conversation and designed for businesses. Pepper communicates with Tally, an inventory tracking robot which rolls down isles, and provides information on what is in stock. Take a listen as the two robots communicate.
02:43 BK: Pepper is saying, pick up your order here, so I’m gonna…
02:46 Pepper: Okay, please give me a few seconds to see if I recognize you.
02:51 BK: So he’s using facial recognition to recognize me.
02:52 Pepper: Hi there, I presume you’re here to pick up here up your online order. Do I have that right?
02:56 BK: Yes, correct. So Pepper is telling me where to pick up my deodorant. So I’ll go next. Is that right?
03:02 KD: Yes.
03:04 Pepper: Okay, now it’s time to meet Tally. Tally is my good friend and also keeps me in the know, especially on things like inventory levels and item costs. Enjoy your time shopping.
03:14 BK: So it seems like SoftBank feels as though retail will be sort of the front lines for robotics?
03:20 KD: It was built for human interaction and to provide information. And so where we see it working is in retail, as the experience that you just saw. Also for greeting, it’s in hospitality, we have it in banks, we’re thinking probably in automotive dealerships. Healthcare, we have a couple of use cases as well. This is just a showcase, really telling a story of how when you connect robots you can get a much more broader experience, and things can be automated, but this is not the end all be all. This is based on the partnership that we have with Tally and Simbe, this is the story that we can tell. Certainly it’s a retail case but there are many other opportunities to use the robot.
04:03 BK: Sure, I think about the vision here. One thing everyone keeps hearing about, I believe some of these stores are going to open very soon, is Amazon’s going to have these automated checkout processes. So people are gonna walk into the store, they’re gonna buy their items, they are gonna walk back out without having spoken to anyone, any human in the store. Do you see yourself in that space then?
04:24 KD: So, the…
04:25 BK: Where like we’re automating the process?
04:27 KD: So what we’re trying to do is with Pepper, you have the ability to automate tasks. What we’re talking about with the cloud infrastructure being built and the data sharing in and then ultimately being pushed out and helping to drive automation, we’re looking at automating entire processes, not just individual tasks. So the example that we have here, there’s a dashboard here, we’ll pretend that it’s on, we’re showcasing what is it that a retail store typically measures and cares about, that’s their gross margin for return on investment or GMROI.
05:00 KD: The more inventory they have in a store, and it’s just sitting there and not selling, the less they’re making money. They’re losing out an opportunity. And so, what we’re showcasing is by increasing basket size by studying and understanding what they have an over-indexed inventory of, we can begin to automate. So the dashboard is typically something that your executives are gonna look at and start to pull levers and say, go do this based on that. What we’re saying is, there’s no longer a need for that human intervention to make those decisions.
05:28 KD: As Tally is going through and recognizing that there’s a certain number of over-index of one of these deodorants, it can then push that information up into the cloud and say, “You know what guys, we need to be promoting this item more than others.” And so, now what you’re going to be seeing on Pepper is based on the automation of, “Hey, we need to sell this. Let’s start promoting that.” You can get into automation of pricing, automation of that promotion. You think about then potentially even warehousing robots getting involved and automate the entire flow of what that is. So, it’s not just, as you have mentioned, what Amazon’s trying to do to automate the check-out portion, we’re trying to bring the decision process and make it a lot more seamless, a lot easier, and more real-time for businesses, so that they can respond and react.
06:16 BK: After Kass Dawson, Global Head of Marketing for SoftBank Robotics, pointed out robotics can help improve profit margins and deliver a higher ROI through IoT automation in the cloud, I take the opportunity to ask, what is probably on quite a few of my listeners’ minds and that’s, will these robots be taking jobs? Should people be concerned that robots will take their jobs, or do you think they’ve already done that?
06:38 KD: People should not be concerned that they’re gonna take their jobs, and there’s multiple reasons for that. One is, with our robot, specifically, with Pepper, the intent is not to steal jobs, it’s to replace tasks. And a lot of the tasks that are being automated are those menial tasks that most employees don’t wanna do anyway. In retail, specifically, if you think about it, I think it’s a quarter of time, of a retail employee is spent on tasks that they just hate doing. And so, that’s what leads to the turnover that’s really high and cost retailers a ton of money trying to make sure that the employees are up-to-date and educated.
07:14 KD: So, the intent is to make sure that what they’re working on is no longer those menial tasks, but things that are higher level, higher touch, and helps them to enjoy their day, so you decrease turnover. Excuse me. The other thing about automation is jobs are going to evolve. There’s lots of reports that say jobs are going to go away, because tasks are going away, but there’s now going to be more jobs around the development and management of that automation process, and making sure that you can do that.
07:46 KD: And so, I believe, the World Economic Forum just did a report a couple of months ago that talked about yes, there’s 50 million jobs that are going to be, quote unquote, replaced, but 75 million jobs are gonna be needed to make sure all of this automation can be done. And so, another piece of what we’re trying to do is education, making sure that people are educated and do so in a way that is no code, low-code, making it a little bit easier for your existing workforce to learn how to manage robots, and AI, and automation.
08:21 BK: My second guest is Richard Margolin, CTO and co-founder of RoboKind, who is doing important things in education by using robots to engage with autistic children. The response from the kids is mind-blowing with an improvement from 2% to 3% engagement with human therapist to 87% engagement with his robot, Milo. What is robots for autism?
08:43 RM: So, robots for autism is, really, it’s our first, and, really, biggest product. And so, we use our facial expressive robot Milo to teach social and emotional skills to children with autism. And so, we’ve, basically, paired the robot with a comprehensive evidence-based curriculum around social and emotional skills and we use the robots to deliver it in public school classrooms. And the thing that’s really cool is, basically, for a lot of kids with autism, almost definitionally, they have trouble with other people. And so, a lot of them don’t do well with human to human therapy. And so, the stats are that kids with autism are only engaging their therapist 2% to 3% of the time, and with our robots, they’re engaging 87.5% of the time. So, that’s, kind of, our big starting point and why this works so well for this population of kids.
09:46 BK: Why is that? Why are robots so much more effective with autistic children at the 87% versus the 2% to 3% compared to humans?
09:56 RM: The robots are really more effective, because they’re predictable. They’re not extraneous bandwidth, they’re consistent. The things we do are things robots are good at in terms of the way we teach. So, it’s perfect repetition, consistency, the robot’s never getting tired, angry, it’s never had a bad day, it’s always ready to go for these kids, and because it’s, kind of, a paired-down version of a human, there’s not all this extraneous, overwhelming bandwidth of information for these kids to have to try and figure out how to filter and process, which they don’t innately know how to do, and that’s, kind of, what a lot of what we’re teaching them in some respects.
10:44 BK: You have another initiative called Robots for STEM. How can a robot… How can a robot help develop computer programming skills?
10:56 RM: So, Robots for STEM is, basically, a coding course that’s tailored for elementary school kids. And what we’ve done is, again, the robot is a cool engagement tool for these kids, and this is every kid, not just special needs. And we’ve developed the curriculum in a way where it’s effectively self-guided. We want teachers involved, but at the elementary school level, we don’t have teachers, generally speaking, that know how to code, or teach coding. So, we’ve developed a course where a teacher, basically, has everything they need to guide these kids through it, rather than to need to be a subject matter expert. And then, with the robot, we also have a simulation of the robot. So, one classroom can have a single robot, and kids can still work on all of their stuff in a browser. And we’ve… It’s much newer. We just launched it this year, but we’re seeing kids, really, kind of, as they work through the course, in second and third grade really able to learn programming and coding concepts, using visual programming.
12:08 RM: And if your “hello world” lesson, for instance, instead of it being black text… White text on a black screen that says, “Hello world” kind of, when I took it, or even just a character on a screen with a pop-up bubble that says it, you get a robot that looks at you, waves, smiles, and says, “Hi”. The first school we ever tested a beta version of this course in, their previous STEM course, which was an elective, had three kids in it, and after they brought in the robot, and kids heard about it, they had to add several more sections because 55 kids wanted to sign up. So, I mean, that’s…
12:49 BK: That’s pretty incredible.
12:50 RM: Yeah. So, when I talk about engagement, it’s not just like a thing, where it’s not just about a kid sees it, and they think it’s cool, so they wanna do some stuff with it. It’s enough of a… It’s different enough, and it’s enough of a draw, that it’s bringing kids who otherwise weren’t considering even looking at trying the stuff out, and it’s bringing them into these classrooms and wanting to really learn it.
13:16 BK: I asked Richard if it was hard to take on education, and he mentioned a few challenges due to public education moving so slowly. Despite this, RoboKind has won quite a few awards. There are too many awards to list here, so, I’ll let Richard tell you about them himself. So, what kind of awards have you won despite this challenge of competing for the press with all these other verticals in technology?
13:40 RM: Wow. We’ve won a lot of awards. The last year, year and a half, has been award heavy, which has been kind of cool. I guess, most recently, I was D Magazine in Dallas CTO for Emerging Technology Award. Last year, we won Launch Festival in San Francisco for best education company. We won Tech Titans’ Inventor of the Year Award. We won several education awards from… I’ve forgotten all of the…
14:39 BK: You’ve won so many awards, you’ve forgotten. [chuckle] That’s good. This is a good sign.
14:39 RM: Yeah. Tech Edvocate, I know gave us awards for both our autism and our STEM programs, and we won Cool Tools Award from EdTech Weekly… Or EdTech Digest last year as well for our autism program. And I know there are others that [chuckle] I’m forgetting. And so, it’s been… The awards are great. They don’t necessarily put robots in classrooms, which, obviously, is what’s important from my perspective, but they’ve gotten a lot of attention and have opened doors. So, maybe, indirectly, they have put some robots in classrooms.
15:27 BK: What would put more robots in classrooms?
15:32 RM: I think, on the autism side, just a willingness to do something different a little faster. We are starting to, really, accelerate in the autism space, but it’s taken several years of finding early adopters who are willing to try out something that was very different than anything else that was out there even after we’ve had several good independent academic studies.
16:05 BK: My third guest is Zack Boniface, a systems engineer in automation delivery at John Deere. To be honest, I was a little surprised to see a full-sized iconic John Deere tractor in the middle of a tech conference, let alone, the robotics section. I pulled Zack aside and picked his brain on why the intersection of agriculture and robotics is crucial for future generations. Well, I think, when people think of robots, they think of the Jetsons and we’ve looked at retail robots today, we… Big robots in the news, or surgical robots. What robotic features will tractors have, exactly?
16:46 ZB: Well, so, the big thing about robotics in John Deere… In our industry, in general, is that we see robotics as a way to help solve labor issues. So, farming is done in very remote locations, often, where there just aren’t people around to do the work, let alone people that are skilled at operating the machinery, or know much about farming. So, if we can use robots and automate a lot of those operations, that will keep a lot of remote acres in production, and continue to produce food.
17:21 BK: Is it easy to find laborers for farming? Farm laborers?
17:26 ZB: No. It’s not. It’s becoming harder and harder. So, a lot of people are moving off of the farm. Farmers are aging and people aren’t coming in to farming to take their place. And like I mentioned earlier, farming takes place in very remote locations where there just isn’t labor available. So, the intelligence that we’re adding to our machines allows less skilled operators to get in the cab and operate our equipment, and allows our farmers to find people that can help them out in their operations.
18:06 BK: How much food does the United States produce compared to other parts of the world? How do we touch into that? Yeah.
18:10 ZB: Well. Yeah. So, the big issue is just, kind of, global population. So, by the year 2050, there’s going to be 10 billion people on earth, and we’re gonna need to grow about 50% more food than we’re growing right now, but the number of farmable acres that are out there, that’s not increasing, in fact, that’s going down. So, we’re gonna do… Have to do a lot more with less in order to feed the world.
18:35 BK: Seeing that John Deere was, in fact, at a tech conference, I thought I would ask Zack about some of the ways data and artificial intelligence can improve the harvesting of crops. How does data inform your tractors? Or what kind of data are you guys using?
18:50 ZB: Yeah. So, all of our machines are connected with 4G LTE modems that are collecting data on the machine as they’re operating, uploading them to the cloud, and then, we offer a number of tools to the customer to access their data. And that data really drives a lot of economic decisions that they’re making. When to plant, what to plant, what nutrients to apply to their fields, all of that data comes through our cloud.
19:20 BK: Tell me how AI helps the industry of agriculture?
19:24 ZB: In several ways. So, we’re highlighting a couple of examples here at CES. One of them is we have technology that we’re embedding into our combines to assess the performance of the harvest quality. So, sometimes, when grain is harvested, some of the plant material gets in with the green. So, we have cameras, strategically, placed on the machine to take images of that green, assess the harvest quality, automatically adjust settings on the combine to bring harvest quality up to where it needs to be.
20:03 BK: How does automation optimize farming?
20:06 ZB: Great question. A good example of that would be our active fill control feature. So, that uses a stereo camera technology to sense the position of the vessel that we’re unloading material into from a harvesting machine. Okay. So, we automate the filling of that machine, so that the operator doesn’t need to use their attention to fill the trailer. That allows them to focus on loading the machine as much as they can, so that they can be as productive as possible.
20:43 BK: Thanks for joining me for the robotics episode. Please support the production of this podcast by leaving me a review in iTunes.
20:53 Speaker 7: This episode was brought to you by Intertrust Technologies. Helping you build data-driven businesses. Go to intertrust.com for more information.
Businesses today depend on data to improve workflows, run more efficiently, and discover new opportunities. Intertrust enables you to build data-driven services tailored to your industry. Securely share data within organization, with partners, and even with competitors.