Sparking Interest: Q&A with with Nimble Robotics founder and CEO, Simon Kalouche

Posted: February 11, 2021
Simon Kalouche
Kalouche

Nimble Robotics founder and CEO Simon Kalouche earned his bachelor’s with honors in mechanical engineering from The Ohio State University in 2014. As an undergraduate, Kalouche was the recipient of Ohio State’s Outstanding Undergraduate Research Award and the first place winner of the Denman Undergraduate Research Forum. He also minored in entrepreneurship and innovation.

Kalouche went on to earn his master’s in robotics from Carnegie Mellon in 2016. Later that year, he began working towards a PhD in Robotics at Stanford University. Kalouche eventually paused his academic work to found Nimble Robotics.

Nimble is reinventing fulfillment with intelligent robots that can pick and pack anything. They already have fleets of robots out in the real world, picking millions of products for some of the world’s largest retailers. Kalouche was recently named one of Forbe’s 30 under 30 in manufacturing and industry for 2021.


 

For those who don’t know, what does Nimble Robotics do?
Nimble is creating intelligent robots, and we’re using those robots to reimagine and reinvent e-commerce fulfillment. If you look at the best Amazon warehouse in the world today, it has a lot of automation but it’s still designed around people: what people can do, where people can go, what’s safe, productive, ergonomic for people. And if you look at what those people are actually doing, they’re doing the picking and packing step – the last manual fulfillment task. When you fulfill an online order, you have to pick the items and pack them into a box. That part is still manual in every warehouse because it’s really hard for robots to do, because there are millions of different products and objects and all of them are different sizes, shapes, weights, textures, stiffness, etc. And what we’re doing at Nimble is teaching robots how to handle all of these objects, all those millions of objects. Once you have robots that can intelligently pick, pack, and handle any object you no longer have to design warehouses around people. You can completely reimagine, reinvent, and redesign warehouses around what robots can do and where robots can go and that unlocks a new kind of robotic warehouse that fulfills online orders an order of magnitude faster and cheaper than the best Amazon warehouse in the world.  In a one sentence summary, we’re building autonomous fulfillment of the future.

Do you repurpose robots or build them all from scratch?
Today we go into warehouses that already exist, and we’ll retrofit those warehouses with picking and packing robot arms. The warehouse infrastructure is already in place, and there’s a station where picking and packing is done, we’ll put a robot at that station. We’re also working on a next-gen robotic warehouse and that’s kind of the future of Nimble.

Did you always know you wanted to be a robotics engineer?
No, the FEH [Fundamentals of Engineering Honors Sequence] program is really what got me into robotics. That was my first exposure to robotics. When I was in high school I was really into architecture, designing buildings and houses and stuff like that. So I always knew I wanted to be technical, and an engineer of some sort, but the FEH program my freshman year, where you had to design a robot to navigate a course, that’s what really sparked my interest in robots. And the reason was because it was so multi-disciplinary, it was not just mechanical design, or not just electrical and circuits, or not just programming and computer science, it was the fusion of all of those things that made it exciting to me.

When you were an undergraduate you worked in a lab with MAE professor Haijun Su. What kind of research did you work on?
I worked with him, and I worked with professor Umit Ozguner out of the ECE [electrical and computer engineering] department who worked on autonomous vehicles.  I worked with him first, and we built this self-balancing bicycle. It used a control moment gyroscope, which is basically a flywheel that spins really fast, and you can rotate the flywheel about an axis and that will induce a moment which you can use to balance a bicycle. So that was the first big research project I did at Ohio State. Then I went to work at NASA’s Jet Propulsion Lab where I started working on a gecko adhesive climbing robot. Aaron Parness, who was my advisor there, he created this synthetic material that basically emulates how geckos’ feet work. Geckos can climb on walls and climb on ceilings, and they use Van der Waal forces to do that. He created this material that has the same properties. And what I did was I used that material on this robot’s feet, and built a robot that could use that material to climb up walls, or up a ceiling. Then I brought that work back to Ohio State when I worked with Dr. Su, and we continued the iterations. We made the robot better, and we made it more capable, and faster and able to do more things. So that’s what my research with Dr. Su was. It was kind of an extension of the work at NASA.

Over your experience through undergraduate, master’s and PhD work, how has robotics education evolved?
Once I went to NASA, my whole view on engineering kind of changed. When I was in school I thought mostly ‘how can I get good grades, pass all my classes, and graduate at the top of my class and get a good job?’ That’s how I thought. Once I went and did an internship, especially with NASA where there’s a lot of freedom and they kind of just throw you in, that kind of just exposed me to the real world of engineering. I was surrounded by insanely smart people, and that put all the equations in my textbooks into context in the real world. All these things came together for me, and that was a pivotal moment for me. Then when I got back, that’s when I started working with Dr. Su, and when I got really interested in grad school. Carnegie Mellon is where I went for my master’s before Stanford. And I’d say Carnegie Mellon is unmatched in robotics. They have a whole institute dedicated to it [robotics]. That was a great experience for me, I learned a lot about robotics and I hope that Ohio State has their own institute for robotics one day. I think that robotics is the future.

When did you make the decision to switch from PhD work to founding your own business?
It was when I started working on imitation learning. Imitation learning is kind of a deep learning algorithm that allows you to train a robot from human demonstrations. The idea is that you can tele-operate or remote control a robot to do tasks that the robot can’t currently do autonomously today. So you take a task like picking up any object, and you tele-operate, you remote control it and you kind of guide it through the motions. And if you do that hundreds or thousands of times, the robot starts to generalize and it’s basing it off all of the demonstrations, so it’s not like it’s repeating any one of the demonstrations you did, but it’s using that experience to infer how it should pick up something that it’s seeing that looks like an object it’s seen before. This framework of imitation learning is what got me really excited about starting a company. I was like we can apply this in the real world today, I don’t have to sit here and write research papers, it’d be a lot more exciting to go find a problem and solve it using this method. The technological inflection point that spawned Nimble was that you now don’t have to wait until you’ve fully automate a task to deploy it in a product. Instead you can have people tele-operate the robots to accomplish that task, provide valuable robotic labor today and as a byproduct of that labor you’re actually collecting training data from the teleoperators in the process. Then you use that training data to incrementally automate this task. If you think of it, you don’t have to sit in a research lab for 10 years working on solving this one problem, then finally it’s ready and then you can deploy it. You can deploy it today, it’s not going to be fully-autonomous, it’s going to be partially tele-operated by humans, but every demonstration the tele-operator gives is another training example, and you’ll continually retrain the robot with all that data and it will continue to get more and more autonomous over time. So it works reliably on day one and it will work reliably on year 10. That’s the framework that got me excited about starting my own business, that this can be applied in the real world today. It allows you to automate tasks that were previously considered ‘un-automatable’ and that unlocks a whole new market of opportunities.

Have you found engineering and business good crossroads that give you an advantage in some sense, in dealing with all the challenges that come with a start-up?
Absolutely. Problem solving and being relentless, relentlessly resourceful, you need that. You need that to start a company. There’s going to be a million challenges, and you don’t have experience with any of them, so you just have to tackle it, break it down, understand what it is and not accept status quo. That’s the other thing. The world works in a certain way, and if you want to disrupt it you can’t take it as it is. You can’t just assume this is how it works, this is how it has to be and this is what I’m going to do. You have to understand why it works that way, break it all down and many times there’s a better way to do things.

What do you find most rewarding about starting your own company?
I would say two things. I would say building a team is really fun. It’s something I had no experience with and I learned a ton along the way. But building a team is so fun, and creating a culture of awesome people, creating an environment where we’re working together toward an awesome goal, that’s one piece. And then accomplishing really hard things together with that team, that’s the other piece that’s most rewarding. Looking back on every step, nothing was easy. Every step of the way is extremely hard, especially for something like robotics. But when you look back on it that’s what I’m proud of. We formed a team and we accomplished some awesome stuff. We did things that seemed impossible, and I like to say to my team ‘we need to make the impossible possible every day.’

Most challenging?
Challenges vary from time to time, they’re always evolving. I think there are challenges around people, and challenges around product, what you end up building and how you deploy it. The challenges around product is that we’re building something state of the art. We’re creating technology that doesn’t exist yet, and we’re applying it in the real world in a completely new market. Then we’re also forming a team, making sure everyone’s aligned and that we have this cohesive vision, and making sure everyone’s happy and working on things they want to work on is also challenging.

Sebastian Thrun and Fei-Fei Li are backers of Nimble, and two big names in robotics. When did you come in contact with them?
I worked with Fei-Fei at Stanford. She was one of my PhD advisors and she’s a legend in AI and computer vision. She created ImageNet among many other things which helped spawn the deep learning revolution in 2011. And I met Sebastian when I was at Stanford. He wasn’t my advisor but we’d chatted a few times and he was always very helpful and insightful, and when I started Nimble I reached out to him to be an advisor, because he’s like the ultimate tech pioneer, he’s like the father of self-driving cars. And he started several of his own companies, so he was a role model to me.

How was the process of transitioning from student/advisor to founder/backer?
I had awesome advisors and professors who were very supportive and that helped a lot. I kind of just told them that this is what I wanted to do and that I was super passionate about it, and they saw the excitement. They encouraged me, said ‘How can we help?’ and a few of them like Silvio Savarese, Fei-Fei Li, and Andy Rachleff who were all my professors at Stanford were actually some of the first angel investors in Nimble. They were just very encouraging and they believed in it. Having that support was very helpful to me. That’s why when I see someone passionate about something, I get very into it. I might not know what it is, but I listen.

What did it mean to you to be named one of Forbe’s 30 under 30?
It is a cool recognition. It felt rewarding and I’m very happy to have gotten it but I don’t think it means that ‘I’ve made it.’ I think until Nimble builds a better automated warehouse than Amazon we haven’t made it. So we still have a lot of work ahead, but it’s all really exciting and we’re looking forward to continuing the journey.

What is the ideal future of Nimble?
We want to build the future of fulfillment. We want to reimagine everything from the inside of the warehouse to your front door, to get you what you want, when you want it, in a faster, cheaper, more environmentally friendly way than Amazon can. There’s still tons of opportunity to do that, and we’re onto something. Stay tuned for the next couple years.

What advice would you give to students who want to work in robotics?
In Robotics, specifically, I would say do hackathons, have side projects, tinker and build stuff. That’s honestly the best way. Find something that’s super interesting or that you’re passionate about and just build it, because that’s the way you learn. You’re not going to learn robotics by reading a textbook. The equations are great, but really to build something you have to tinker with it, you have to wire up the Arduino, you have to create a motor controller, hook up a camera, do some computer vision. And open source has made it so easy to just dive into. So I would just encourage people to build stuff. That’s the best way to learn in my opinion, by doing.

What advice would you give to students who want to start their own company?
I would say one of the most important things is you have to be so passionate about whatever you’re building, and truly believe in it. There’s the whole missionary vs. mercenary debate. If you’re doing this for the money there’s going to be a million things that are hard, and a lot of tough days where things aren’t working. And if you’re doing it for the money it’s going to all break and you’re going to give up at some point. If you’re doing it because you believe in this mission and because you’re passionate, that’s what’s going to keep you going through the hard times. Ask any successful founder of a huge company. There was no legendary company that was built easily. You have to be passionate and you have to be relentless. Just never give up. It’s a very challenging but very rewarding experience and the best part to me is the endless opportunities to learn every day. As a first-time founder, everyday I’m always learning new things. When I started Nimble I didn’t know how to pitch investors, I wasn’t trained in selling things, so I had to learn to sell and raise money. But you get to learn all these things, how to manage people, how to manage projects, and that’s the fun part to me. It’s just perpetual learning. Everyday you’re thrown new things without a clear path of how to do things and you’re forced to problem solve, you have to figure out how to get good at those things. If that all sounds good to you I think you’re probably a good fit to start something.