Tell us a bit about you and your venture! What was your inspiration for creating your venture?
I started building robots with perception capacities in 2010 when I was an undergraduate student. By the end of the 4 years, I had competed and had created several award-winning technologies in the robotics space. After graduation, I started graduate research in human-machine interaction at Autonomy Lab, led by Prof. Richard T. Vaughan at Simon Fraser University (SFU). My graduate research was all related to the design and implementation of visual perception methods for robots. I also completed an business graduate program focusing on science and technology commercialization in SFU, where I implemented the first version of the business plan for InspiRED Robotics with the help of Prof. Elicia Maine.
InspiRED Robotics is dedicated to building deep neural network based machine vision solution for embedded devices. It helps drones, smart TVs, toys and other consumer electronics products to understand humans’ gestures and behaviours, so that they can interact with human more autonomously and efficiently.
Understanding human behaviours is the key for intelligent machine systems. And vision is the most important perception method for humans to understand each other. In Autonomy Lab and Vision and Media Lab in SFU, where the co-founding team of InspiRED Robotics comes from, researchers has been working on machine vision techniques to help machines to understand humans and interact with humans autonomously for more than 10 years. Now it is the right time to commercialize the technology since we finally have processors and software algorithms good enough to handle the task.
What problem is your venture solving? Why did you choose to tackle this market?
When we talk about machine vision or artificial intelligence in general, we usually think about those super computers in giant boxes running on the cloud. However, the super computers are not always available because of the price, size and power consumption, especially for mobile devices with embedded systems. And there are other issues for cloud computing including transmission delay and data privacy. Most of existing solutions cannot solve the problem yet. We chose to tackle this market since we would like to make deep learning based machine vision technique more accessible, not only for the super computers but also for the mobile embedded platforms.
What are some of your venture’s biggest milestones?
Through one of the networking events of NextAI and the Ontario government, we got connected with an important strategic partner in China to help us get into the Chinese market. The first customer is always the hardest to get. After we established the partnership with our Chinese partner, we got another two B2B customers right after we graduated from the NextAI program.
When you first started, what were your biggest hurdles in building your venture?
When we first started, most of the team members come with a technical background. Although I have a business degree myself, there was still a lot to learn to start a real business. We spent a long time to figure our the related policies on business registration, accounting, tax, and other topics. Another challenge is to get the first customer. We sent a lot of cold emails and did a lot of cold calls as well just to figure out whether the market exists and what the customers actually want. According to our experience, it is usually easier if you can get a warm introduction to your potential customers by someone they already knew, than simply cold calling them.
How do you believe technology will impact your industry over the next decade?
Machines equipped with visual perception capabilities and powered by the artificial intelligence will serve the humans better. This will not only happen to the consumer electronics, but also to the industrial machines. The deep learning technology will help solve the problems that was hard to solve before, for example to detect certain objects and recognize gestures in very dynamic and complicated environment reliably.
What piece of advice would you give to someone who wants to start a company in your industry?
Although technology itself is certainly crucial for an AI startup, my suggestion would be to put enough focus on real-life applications and use-cases. Just to make sure that the needs and the markets actually exist before digging too deep into the technology development.
What are 3 books, blogs or newsletters you recommend for entrepreneurs looking to make an impact in your industry?
- Elon Musk：Tesla, SpaceX, and the Quest for a Fantastic Future
- IEEE Spectrum Robotics: https://spectrum.ieee.org/robotics
- The Robot Report: https://www.therobotreport.com/map/
What would you say are the most important skills needed to be a successful entrepreneur?
- Always working hard: I mean literally putting in 80- to 100-hour weeks every week.
- Making right decision with limited information: I learned this from Reza Satchu who is part of Next Canada and also a very successful entrepreneur.
- Willingness to learn and learn fast: the world is changing fast and as an entrepreneur you have to move faster by learning;
Who is one person that has tremendously helped you through your time in NextAI? How did they help you?
Annick Dufort, the program manager of NextAI, has been so helpful during the whole process. As one of the entrepreneurs in the first cohort of NextAI, I would say the program is very well organized and I can see that Annick and the Next Canada team have put lots of effort in it. Annick not only helped on “small things” such as working space arrangement, but also helped making connection with important business partners for us. When we attended events together, she was always happy to make introductions for us to other attendees. I would like to say thank you to Annick and the Next Canada team.
Applications for the 2018 NextAI cohort are open! Click here to apply.