.lumen: Meet the team empowering millions of people who are blind through AI and machine learning models – Cisco Blogs

Now that the Cisco Global Problem Solver Challenge 2021 winners have been officially announced, we are excited for you to learn more about each winning team and the story behind each innovation. The Cisco Global Problem Solver Challenge is an annual competition that awards cash prizes to early-stage tech entrepreneurs solving the world’s toughest problems. Now in its fifth year, the competition awarded its largest prize pool ever, $1 million USD, to 20 winning teams from around the world.

Last December, we announced that the 2021 Challenge would include a special Ethical Artificial Intelligence (AI) Prize.  The $50,000 USD Ethical AI Prize was awarded to a startup designing AI in an inherently ethical manner, so that the solution addresses social, environmental, or technological challenges. This award is being offered by Cisco’s Emerging Technologies and Incubation (ET&I) Group in the spirit of our purpose to Power an Inclusive Future for All. The partnership between the Cisco Global Problem Solver Challenge and ET&I is symbolic of our commitment and investment in the corporate social responsibility space and our passion to be at the forefront of innovation.

This year’s Ethical AI winning solution, .lumen, is focused on the same goals as our ET&I team, a commitment to and investment in the artificial intelligence and corporate social responsibility spaces. .lumen builds technological glasses that replicate the main advantages of a guide dog, and then some, in a package that can easily be replicated, maintained, and used. The solution’s AI and machine learning models will provide enhanced mobility and opportunities for people who are blind.

Our ET&I team sat down with .lumen’s Cornel Amariei, CEO, to discuss their innovative product, their inspiration, and how winning the prize will impact their business.

What problem is your technology solution trying to solve?

Cornel: Right now, if you look at the world, there are 40 million blind people. This number will increase to 100 million by 2050. Despite all the technological advancements – you only have solutions allowing a blind person to read text or use a smartphone or laptop. And, you had the exact same two solutions for mobility and orientation for thousands of years – the white cane and the guide dog.

Now, the guide dog is very interesting. Its features are unanimously seen as helpful, but it has a few drawbacks. It costs between $30-$60K to train a single guide dog – and that’s not even the biggest problem. By far, the biggest problem is that a blind person can’t easily take care of a guide dog. And because of this, there are only 28,000 guide dogs to 40 million blind people. So the most advanced mobility solution, and the most useful one, is not a scalable one.

We realized that today, we could mimic the main functionality of a guide dog without the drawbacks and create a system for the other 39,972,000 people.

Can you explain how the solution works?

Cornel: I’ll start with an analogy. If you have a guide dog, you can ask it to take you to the door, to an empty seat, or back home if it knows the route. It will do so by pulling your hand, moving you away from obstacles as you go. You can ask the .lumen glasses the same thing, using a voice command. But rather than pulling your hand, they will pull your head.

We have a patented system, which we evaluated with over 200 blind individuals. It very intuitively uses haptic and auditory feedback to guide a person very precisely – avoiding obstacles, avoiding dangerous situations to keep you on track. And it’s as easy as someone pulling your hand in the direction you need to travel.

On a more technical level, we basically took technology from autonomous driving and robotics, and we scaled it down to something you can wear on your head. It’s an immense trick of engineering, because you have so much space and available weight for sensors, batteries, and computers in an autonomous driving car.

In our solution, we only have a few hundred grams or roughly 20 ounces. The system has five cameras. We use artificial intelligence to understand the situation, to understand how you can interact with the world, and where it is safe to travel and where it is not.

We, as visually capable individuals, take these things for granted without realizing how complex they are. Our headset operates in our custom hardware and software system – and it’s an incredible success that we’re getting close to what a guide dog can do.

What inspired you to develop this solution? 

Cornel: The company is just under one and a half years old – so it’s a pandemic startup. I left my job as the Head of Innovation at a very large automotive corporation to start .lumen. My motivation and inspiration actually came from my family. I was born in a family where both of my parents have locomotive disabilities, and my sister has cerebral disabilities. So I grew up being the only member in the family that does not have a disability. This made me very sensitive to the experiences of persons with disabilities and really made me realize how assistive technology can help – and how little is being done.

Then, being an engineer, scientist, manager, and someone who has built other companies, along with my family experience, allowed me to put it all together. And day-by-day, we’re working with over 200 blind individuals to develop and test these products. We’re very fortunate that way.

How will winning a prize in the Cisco Global Problem Solver Challenge help you advance your business?

Cornel: First of all, what we’re trying to do at .lumen is incredibly hard. And we need all the support in the world. Any type of support, especially what we’re getting now from Cisco – the prize, the mentorship, and everything we’re receiving – definitely helps us focus on what we have to do and improve our testing efforts. With everything we do, we test, and most of the money goes there.

Do you know what you will use the prize money for specifically?

Cornel: We already allocated the budget for it – even before we had the money – to product manufacturing. We are beginning manufacturing soon, starting with 200 units, and have a global expansion plan next year. And in 2022, we’ll make an additional 10,000 units.

How has the global pandemic impacted your work?

Cornel: Thankfully, there are many things you can do remotely. Quite many more things than I would have said would have been possible two years ago. Now, COVID has shown me how many things can be remote.

But there are things which you cannot. For example, you can develop hardware remotely, but it’s not easy. And you can’t really prototype it that way. I expected that to be the biggest challenge, but we took note of that. The biggest challenge by far was testing, because we must get qualitative and quantitative feedback from over 200 people. They have to get our devices. We have to track how they interact with them, etc.

In order to test in-person during COVID, during lockdowns – forgetting about the tremendous legal hassles, the risk that we took was incredible. Under very strictly controlled conditions, for every hour of testing, we had over 100 pages of procedures. We have hundreds if not thousands of pages of procedures as a result.

Some of our testers had COVID. They didn’t know it, and we found out after the fact. And still, we never got it. We basically created a sterile hospital environment in our office. It was the only way we could test the system.

Why did you decide to start your own social enterprise versus going to work for a company?

Cornel: I was a university student back in 2014 when we first got the idea for .lumen. We actually built some prototypes back then to test them with blind individuals, and we did so. But it was quite clear to us that the technology was not where it needed to be. So knowing that we couldn’t deliver it then, we took it upon ourselves to wait for the moment technology would catch up with our vision and our idea – and that’s the moment we’re going to start. That moment was 2020. We anticipated that in 2019 and said, “Ok, if this is the shot. The technology is here – let’s start.”

My cofounders and I put in our money and started the company. One year ago, we were nine people. Today, we’re a team of 40 passionate individuals. Next year, we’re projected to expand to 100.

What advice do you have for other social entrepreneurs? 

Cornel: There’s one thing we learned from .lumen, which is: make meaning, not money. The problem in the world, not just with startups, is that people prioritize money, and then find a solution. Afterward, they find a problem for which the solution works. It should be the other way. You see a problem and really understand it, and then you come up with a solution that fits. Its only purpose should be to help. The moment you really make that work; the moment you really help – that’s when you make money.

That’s what really worked for us because we created .lumen with only one purpose: helping.

Stay tuned for more articles in our blog series, featuring interviews with every Cisco Global Problem Solver Challenge 2021 winning team! Finally, if you are interested in partnering with Cisco to innovate for the future, please consider joining Emerging Technologies & incubation’s Design Partner community.

Share: