Thought Leaders in Artificial Intelligence: Edge Impulse CEO Zach Shelby (Part 5)

Sramana Mitra: I enjoy listening to you and hearing what you have been doing. I keep my eye out for finding platform companies. You may want to look up my writing on the Platform-as-a-Service. I think you will enjoy that. It’s an area that we are covering extensively.

Switching gears, what are some open problems that see that you would like some new startups to go pick up using your platform? From where you sit, you are developing the platform and you are looking for developers to develop applications. If you could have your wish, what are some apps that you would like to see entrepreneurs take on using your platform?

Zach Shelby: We do a lot of work for tech for societal good. It’s part of the culture in our company. We invest time in nature conservation, education, and health. We want to see this technology being used to solve societal problems. In our tech for good work, we run across this problem. There are no AI-capable cameras. These are cameras where you can put machine learning in them and use them for nature conservation applications and other applications too. That area is screaming for a good open camera platform.

All the stuff available today is like the Bushnell camera trap. It just logs stuff on an SD card or maybe sends it to a web server. What people in nature conservation need is something that can classify what is happening. This includes classifying what is happening or classifying a particular animal right on the device.

That is the kind of thing that our technology can enable, but we don’t know what the hardware products are. We don’t use devices that can be used in that way or are open enough to be used in that way.

Sramana Mitra: Somebody should take that on and build something.

Zach Shelby: Yes. Someone should build an ML-powered camera that is easy to put up with a battery, for example, and put it somewhere where it can detect a specific animal. There are a lot of industrial applications in this kind of thing too. We get asked all the time by automobile manufacturers where they have specific computer vision problems.

For example, they would say, “I want to detect when this bolt is going on wrong. It’s the wrong bolt or it’s being assembled wrong. I want to be able to detect that with a camera that is trained on some example images and be told when that happens.” These are simple applications in manufacturing where these smart cameras can be used.

I think that is an interesting area. Someone can go make an easily programmable ML camera for manufacture to solve specific problems. It can be clicking a button twenty times to show an image example of good or bad. You can train that model and put it back on the device and set an alarm when it detects the bad. If you can nail that, then the sky’s the limit. 

Sramana Mitra: That is super cool. I am thrilled to be talking to you. I love what you are doing. Thank you for your time.

This segment is part 5 in the series : Thought Leaders in Artificial Intelligence: Edge Impulse CEO Zach Shelby
1 2 3 4 5