On the moment I'm quite excited because it's my first opportunity since quite some time to search for a new position! I'm looking for work in which I can use my expertise in machine learning and/or embedded engineering for the greater good.
👯 I've worked with my team at Crownstone on an open-source switch/dimmer which is better then anything before! 😄 Check it out at https://crownstone.rocks. Also check out the firmware itself, which is called bluenet.
What's cool about it?
- Again, it's open-source! You own your smart home. Period.
- This hardware has support for Bluetooth scanning. This means you can respond to someone's presence not only if they carry a smartphone, but also when they wear a wearable. Asset tracking / lost and find stuff is a cool extra.
- It's sampling current and voltage pretty fast (5 kHz). This means you can go bananas on analyzing individual sine waves to check out what's plugged into your outlet or detect anomalies.
- We're in the process of supporting microapps (small Arduino programs on the hardware). This makes it customizable by anyone who can code a little bit.
- The hardware forms a mesh, Bluetooth Mesh. In combination with Bluetooth Low Energy (BLE) it's possible to hook up everything that speaks BLE. Next, implement "BLE over Bluetooth Mesh".
🥷 I've been working with computers since I was 10. It started with solving puzzles that I otherwise couldn't solve in a popular science magazine (called Kijk) using QBasic. I also played extensively with the hardware settings too (remember "extended memory"!). One time I broke my father's computer. At the computer shop they had difficulties to get it back in order. To my surprise my father was proud. I've been a tinkerer ever since.
- Please, if I reverse engineer your hardware or report bugs in your software, don't shoot the messenger!
- I've seen so much software that is like "Swiss cheese" (in Dutch "gatenkaas") that I won't even get in contact. If I report, it's out of respect for the quality of your work.
- If you work in customer support, please address in your organisation the need for a standard procedure with respect to vulnerabilities. If you don't have the budget for a bug bounty program, respond fast, with respect for my time, and with respect for the responsible disclosure.
🔭 I’ve been working on nonparametric Bayesian methods for point clouds. Hopefully that will in the end result in some serendipitous insights in non-related math or algorithms.
Some other personal interests:
- Consensus algorithms (back from my days working on self-organizing wireless sensor networks). Gossiping algorithms, voting mechanisms, cognitive radio, etc.
- Autonomous robots (worked on evolutionary methods, in particular artificial gene regulatory networks, and sensor fusion).
Wants to learn more about:
- State-of-the-art blockchain technology, e.g. Substrate. Of course, tying into IoT would be fun.
- Deep learning (who not?), e.g. incorporating something temporal like polychronization (Izhikevich). Can we build something that has those beta, gamma, etc. waves? Don't even care if that behaviour is barely functional.
- Homomorphic encryption. It's the future of privacy.
Wants to help build/create/foster companies that want to:
- Aid the electrification of particular difficult sectors/applications where it has nevertheless a large impact.
- Implement wireless charging at a distance of a couple of meters. Charging is (and cables are) the most important barrier to widespread adoption of chips in everything. Backscatter tech is awesome. Without this it's still the stone age.
- Grow lab-grown meat. Without the use of animals, meat is an excellent product. Same is true for fish, eggs, milk, butter, and cheese.
- Build energy tech at the physical periphery of Europe to foster the local economy and move the entire continent faster on the path of sustainability.
- Grow sea farms. Requires extensive robotization to make aquaculture economically efficient. Might also require getting sunlight into deeper sea levels. Big advantage is that the sea is 3D. It's a huge volume to grow crops in.
- Build embodied artificial intelligent systems (robots, self-driving cars, drones). We won't have enough hands in the future.