Drivers Ed for Self Driving: an Interview with Dr. Helen Loeb, CEO of Jitsik, LLC

Dr. Loeb leads the R&D and business development of an early stage startup called JITSIK LLC which develops tools and methods to prepare drivers to Advanced Driver Assistance Systems (ADAS) and semi-autonomous or self-driving vehicles. Jitsik’s goal is to promote education so drivers know what to expect when they get on the road. Jitsik develops Immersive Virtual Reality Driving Simulators to safely expose drivers to driving scenarios and teach controls and automated features. Even with vehicles of level 2 (L2) or level 3 (L3) autonomy, drivers need to monitor the road at all times. They need to know if and when hands-free steering is a safe option for driving. “Hands free” driving is already on the road in a number of commercial vehicles. Jitsik LLC will be your coach to vehicle automation. 

Mike:  Hello Helen! Thank you for being on this interview. So, I’ll just get started, Helen with our first question.  So, could you tell us a little bit about your background please?

Helen:  Sure. So I graduated in 1996 with a PhD in robotics from the University of Bordeaux in France. I hold a Master of Science in Aerospace Engineering from Sup’Aero, number 1 school in France for Aerospace Engineering, and a BSc in Maths and Physics. Before working on my PhD, I worked for a couple of years in the software and aeronautics industry, namely Airbus. Then, as part of my PhD, I designed an autonomous boat for the port of Bordeaux. So from 1994 to 1996, I was already working with GPS and thinking about autonomous navigation. So that’s my training. 

Mike:  You designed an autonomous pod for a what? I’m sorry, I didn’t get that.

Helen:  Oh, autonomous boat. Autonomous boat for the Port of Bordeaux Authority.

Mike:  Okay.

Helen:  A small boat.

Mike:  Okay. All right. I just want to clarify that for viewers and listeners.

Helen:  So I received my PhD then. And I moved with my family to the US. At this point, it was a bit of a challenge to find an academic position in Philadelphia, where my husband worked. So I took a position in software engineering in a small company. For about 10 years, I worked in graphic processing and software engineering, designing scanners and character recognition software. And eventually at some point, I wanted to come back to academia. I needed something more challenging, so I taught at Temple University, in Engineering. And I worked at Villanova University for like two years.

I developed a robotics lab there: the Dynamic Systems Laboratory. Interestingly I was also working on autonomous boats at Villanova. I enjoyed working with students and helping them with hands on electronics, mechanical engineering, and the programming work that goes with it. When our grant ran out, I had to find something else. I had a good friend who had this lab at The Children’s Hospital of Philadelphia. She created the lab at the intersection of medicine and engineering to research traffic safety. The lab is called the Center for Injury Research and Prevention.

She knew I was an engineer. We were close friends. She said, “We could use your expertise analyzing data for driving safety”. So I said, “Sure, I’ll come see how I can help” and I started working over there in driving simulation, analyzing data that, you know, my colleagues were collecting. And I have been working in that field, driving safety, for the past seven years. And even though my path is far from linear, especially for someone working in academia, my training could not have been better for what I do now. My Masters and PhD in robotics, one year at in Boston University and MIT working on neural networks. This is what you call now machine learning, but we called it neural networks back then. And my experience designing software in industry taught me a great deal as well, very practical skills about what works and what does not. I always enjoy learning new things.

 It is Interesting how robotics became such a hot topic, with drones and everything. And also with self-driving. Academic labs have been developing autonomous robots for years, but something was still missing, and suddenly researchers decided that they are going to try this technique of, you know, machine learning for autonomous driving and it worked really, really, well. That was in 2009. Since then, things have gone like crazy and we are where we are now, where you no longer need to steer on highways, in commercial cars. It is mind boggling how we got here so quickly. I mean the DARPA Urban Challenge was in 2007 and autonomous driving was still in its infancy then.

Mike:  That’s amazing. And does that lead right into where you are now?

Helen:  Yes, it is pretty incredible, it all makes perfect sense now. My robotics training, the neural networks stint, software engineering development, I could not have been better prepared for what I do now.

Mike:  Okay. That’s incredible. And while you’re doing this now, you created a side project but it’s really not a side project. Tell us about what else you’re doing.

Helen:  Okay. So I have been working with that driving simulator that we have at the Children’s Hospital of Philadelphia. We run a lot of studies to study driving behaviors. We bring in people and have a strong research program in teen driving. So, my son, Jonathan was studying engineering at Drexel University and he visited me at my lab. And he told me, “You know what? We could do this in a much more efficient and affordable way.” Jon loved games and he was playing already with Virtual Reality headsets.

By itself, it was an interesting idea. Smaller, more affordable driving simulators. Sure, we could do this in a more efficient way. But then with my robotics background and interest in safety, I was following self-driving development very closely. It was 2015 and Tesla had just released its very first version of the Autopilot. Part of me was thinking, well people are going to need some training to this technology. It is kind of a paradox because if it’s self-driving, why do you need training? Suddenly we had videos online of people terrified of self-driving. There were all these polls of people not trusting the technology. So we started thinking, well maybe the driving simulation can be that tool to safely introduce people to this technology. Some people will be comfortable from the start but some other people, like the older folks, they might need to be eased into the technology. So, if we can develop something that can be deployed massively, then maybe we have something that can really help people.

My son Jonathan was going to do his third co-op for Drexel University. This is a full time 6 months internship that Drexel requires from students. And the best thing is their entrepreneurship co-op. Students can pitch an idea. If Drexel likes it, they will fund the student so they work on their own startup. This what Jon did. He pitched the idea of that VR low cost portable driving simulator and got funding and support from Drexel University. So Jitsik was incubated at Baiada Close School of Entrepreneurship. This was January 2017. The company itself was funded in July 2017; a patent filed in November 2017. I’ve been helping my son the entire time. I still work for CHOP, but I also work part-time for Jitsik. And we have right now a number of students who are coming to help in engineering, in business, in mechanical engineering software, so that’s what we are doing right now.

Mike:  That’s great. So, you have a leadership knack, and basically started out with engineering and technology. And then as you went along in your career, you saw the simulator at Children’s Hospital, which I didn’t even know they had one for driving. You noticed how that was working and then your son saw an opportunity to trade something that might be more efficient and service different needs and cheaper. So, you spotted a niche there and then started developing your own.

Helen:  Right. Absolutely. Jon first had the idea of using VR headsets for driving simulation. Then we saw the need in education in Advanced Driver Assistance Systems and self-driving technology. We started to think of business models, how to introduce the simulators in driving schools, maybe dealerships so that you can try advanced driving features, you know, from the showroom before going on the road. So you can try features in a safe way.

Our simulators are very portable. We can bring them to community centers, to wherever. Easy to set up and you can immerse people in technology.

Mike:  So, take us through it – to explain how it works? Someone sits in the seat. And in the driving simulator, you sit in it, then… what happens? 

Helen:  So yes, you have a seat. And then of course a steering wheel and set of pedals.  And you wear a headset and you look through it. Right now, we are using the HTC Vive, a virtual reality headset. So, once you put on the headset, you are, totally immersed into a car and driving scene. So right now, we simulate a Tesla. So, when you sit and wear the headset, the steering wheel is that of a Tesla, and the dashboard is that of a Tesla. And you can use the steering wheel and pedals to drive in a highway, in a city. We have a number of scenarios. And it all looks pretty realistic.

So we have a city scenario. If you drive in the city, you can of course drive manually, make turns, stop at stop signs, the way you would typically drive. But in addition, you can try various Advanced Driver Assistance Systems features, like Cruise Control, Lane Keeping or self-steering. And if you engage the autopilot, like on the highway, the Tesla will steer for you just the way it works in real life.

And you can physically let go of the steering wheel and you’ll see through your headset that the steering wheel is rotating by itself. In addition to that, if you remove the headset, you’ll see that the real steering wheel is also rotating by itself. So you get a feeling of what would happen on the road. And we can teach you with the simulator when it is safe to engage the autopilot, and when and how you should return to driving the car manually. We can teach you how to pay attention even with the autopilot engaged, because you are still supposed to monitor the road, at all times.

So, what we want to show with this driving simulator is what it feels like to actually let go of the steering wheel and stress how you are still required to monitor the road and take over if needed. We want to help people relate to this idea of letting go. It is a simple idea but a psychological milestone that some people will need help with.

Mike:  Yes. I tried it. I felt like I was stepping into a video game. That’s what it felt like.

Helen:  Right, but we are adding some patented features that will make it feel totally real.

Mike:  Yes, it was… You stepped right in and everything was very cool. So that’s a whole lot of your background, and history and what you can do with that kind of training. So, with that in mind… for students, and you’re around a lot of students in University City; what academic suggestions would you have for them if they’re interested in getting involved with what you’re doing? What would you tell a freshman starting out?

Helen:  If people want to get involved, in this self-driving revolution, you should know that there are going to be opportunities just about everywhere. We know that there’s legislation that we need to write. We know that there are going to be tons of business opportunities. Obviously, I’ve been more interested in the technology side of things. So, if you are interested in the technology and you want to be involved, I would say, first of all, to read the tech news, read as much as you can because that’s a very simple step. And there are several things that you can do as a student; you can look for opportunities to go to conferences. There is, for example, a conference in Princeton next week and there are more and more opportunities to go and talk to the people who work in the industry.

And right now, we have a thriving economy so there are lots of opportunities to help. One practical thing to do, for students who are enrolled in school, is to look for ways to volunteer, even if they don’t get paid. With the startup, I have some students who just want to volunteer so that they get into the industry, learn about the tech and get their name posted as well. By writing blogs, for example, or emailing folks on forums, they get involved. It’s another step. Once you start, you know, the next step follows. So yes, do as much as you can I would say, and volunteering is a good way to get at the door when you don’t necessarily have a background.

Mike:  Now that’s great because we always focus on what classes, programming classes, etc. I’ve even been told students should take risk analysis classes. We know a lot of that but networking, yes for sure..

Helen:  Networking is key. But yes, while in school, look for the right classes to take. As far as engineering goes, there’s machine learning, statistics, robotics. You should look at the university you attend, and see which Professors are involved in self-driving research. Best class at UPenn is embedded systems with Rahul Mangharam. He is involved with self-driving research with Intel. If you study at Drexel, we have a machine learning project with Chris Yang and Santi Ontanon. You get to do some big data analysis, which is so important in the industry just now. Whether I work at CHOP or Jitsik, I always take students so there are opportunities there too. But whatever university or college you attend, you should always be proactive and do that little step that will bring you closer to the field.

Mike:  Okay, you’re right. There are no mysteries. Get involved and ask questions and that will lead to more opportunities. So, with that in mind, let’s talk about job skills will be in demand? There’s a lot of new technology happening. There’s a lot of research happening. What skills would someone who’s going to have a job in five years from now be developing?

Helen:  I guess we always need the basic, engineering skills, electronics, mechanical design, these types of skills that I acquired studying space engineering, I can now apply to self-driving. The new skill is machine learning, because this is what is fueling, the algorithm, the industry. We call it big data. You need to learn how to process lots and lots of data. So, things like, statistics, yes are useful. Like I said, machine learning, data analysis, image processing, signal processing, predictive analytics. There are also opportunities for people who are interested in things like game design. Because driving simulation is now used to teach cars how to drive. This is how the algorithms learn these days.

So, interestingly, people who come from game design now have tools like Unity 3D or Unreal, which can help with car design and scenario development. So there is right now this convergence of tools that were typically made to develop video games. But these tools are increasingly used by the auto manufacturers to design cars and visualize things. The industry is starting to use virtual and augmented reality to actually see what these cars look like and what they feel like. It’s almost like game design is high jacking CAD design. So, that’s an interesting thing as well.

Mike:  That makes sense. It’s already built. It already serves the purpose.

Helen:  Because those tools are very easy to learn and very easy to deploy, they might be easier than the typical CAD programs, Solidworks and that type of thing. You can create 3D objects, driving scenarios and play them. Things become more interactive and you can visualize everything with VR and AR.

Mike:  Very interesting. Yes, it’s great because if student know that going in early, they have a little bit of a head start.

Helen:  Game designs is interestingly like VR. It’s new on lot of campuses. Drexel just started a year ago. As an example, my nephew did an internship with me last summer at the JITSIK startup. He came from France. He was exposed to Unity 3D. He was exposed to the Virtual Reality Development that we are doing. He worked with our HTC Vive Headset. When he returned to France, he was hired to work to help build the next generation of flight simulators for Airbus Industries. The loop is closed…

Mike:  That’s great success story. That’s exactly what these kids… I call them kids but they’re going to have such a head start when they graduate because I interview a lot of recruiters and you probably know recruiters too. They would love to see not just your resume but what have you done.

Helen:  Absolutely and so this kid, my nephew, David, had this on his resume, while no one else had it. His school did not offer that type of curriculum just yet. So when he returned to France and put it on his resume, he stood out from the lot. No one else had that experience just yet. So his current job is now to test all the Virtual and Augmented Reality headsets for the Aerospace industry in Toulouse, for Airbus and the like.

Mike:  That’s great. I’m out of questions and we’re out of time but do you have anything you want to add regarding student advice, or we can just move on to how do we learn more about Jitsik?

Helen:  Yes. To learn more about Jitsik, I would say check, of course our website we have a bit of info there. Check the social networks. We have Twitter, Facebook, Linkedin Instagram. We are writing more and more blogs so, you can catch up with us this way. You can also find us at all these big conferences, the Consumer Electronics Show in Las Vegas, we had a booth there last year. We’ll be in Princeton for the smartdriving cars summit next week. We’ll be in Orlando at the Automated Vehicle Symposium in July. Of course, we were at the Venture Café with you, Mike, right here in Philadelphia. But the best thing at this point if a student is interested is to shoot me an email. I love to work with students. I welcome anybody who has the passion and wants to get involved.

Mike:  Well thanks for the fascinating interview Helen!

Helen:  Thanks Mike!