Mark Theunissen

WatchBot: A.I. powered video alerts

This is the story of how we used cutting-edge artificial intelligence techniques to reduce and prevent crime, in an area plagued with constant incidents.

I created the WatchBot A.I. software to detect people in our live camera feeds, and alert to user’s phones with Telegram. The system has been operating in production in Cape Town since 2018, and in this time has been used to apprehend many suspects, both before and after actual crimes were committed.

The application, written in Go, runs at the edge using a Raspberry Pi device. The A.I. model is executed on an Intel Movidius Neural Compute Stick.

Vole: peer-to-peer social web

Vole is a web-based application for sharing words, pictures and videos with others, without a central server. It’s built on the power of BitTorrent, Go and Ember.js.

Vole was founded by myself and my friend Aaron Forsander in May 2013. This was a time shortly after the Arab Spring, when social media was mostly believed to be a force for positive change in the world, and we were all very concerned with NSA spying and privacy.

We created a tool that could help evade government censorship, building on top of some great new BitTorrent technology.

Training Wheels: open source for education

The year is 2012, and you want to teach a group of students some aspect of web development. You’ve got them in a session for 1 day, and you have no control over which laptops they’ll be using. How do you ensure that each student begins the class with a preconfigured environment, ready with your course material?

We created Training Wheels at Four Kitchens, to solve this problem for our own training sessions, and we open sourced it in the hope that others might find it useful, too.

I was the lead backend engineer on the project.

LPR RasPi: licence plate recognition at the edge

In 2016 I worked with a non-profit, community-based security group in Cape Town, developing alternative technical solutions for license plate recognition. Their existing system transmitted video feeds over wireless links, directly into a central control room where all the LPR computation was done. The architecture had numerous problems, but the biggest was that the wireless signal was subject to interference, and regularly dropped off.

I built this proof of concept around the OpenALPR project, to demonstrate LPR running on a Raspberry Pi. We wanted to trial an alternative approach by running “at the edge”, which means deploying a low cost device to do the computation at each camera site.