It’s the end of 2023, and I’m taking stock of the things I’m spending my time on.
The last time I had a legible “job” was working at Stripe, from 2013-2019. Back then it was simple to answer “what do you do?”. These days, less so. That’s a good thing (it’s a great thing), but it can also sometimes be awkward. Maybe writing some things down will help.
I live on Galiano Island, a small community between Vancouver and Victoria in the Salish Sea. After a decade of working from here but mostly for Silicon Valley, I’ve been enjoying working to directly benefit the place where I live.
My longest-running project along those lines is GAIA, the Galiano Association for Internet Access. Galiano is generally quite poorly served by the big telcos. There is fiber to the island, but it’s not available for residential connections. I started a non-profit which leases a couple of fiber lines from the telco, and then distributes the bandwidth from them to a few hundred of the ~1000 homes on the island, using the only practical approach: microwave radios mounted near the top of 100’ or taller Douglas Fir trees. (The island is hilly, densely treed, sparsely populated, and culturally resistant to communications towers. The only way to get line of sight between any two buildings on the island is to go to the top of the tallest tree near each side). We typically achieve about 25Mbps down, which is plenty for a zoom call; we can model our peak usage pretty accurately with (number of subscribers) * 3Mbps. Still, it’s a lot of ongoing work to keep the network robust against radio interference, the weather, tree growth and death, power outages, and so on.
The biggest technological advance during this project was finding a LiDAR dataset from which I could derive accurate heights for every individual tree on the island, and then model which ones could see each other. This avoided a lot of guesswork, drone flights, and expensive trial and error hauling equipment up the wrong trees. It’s also spurred an interest in LiDAR data acquisition and related areas like DGPS-based compasses. I haven’t spent that much time on this stuff yet but hope to do more.
A smaller, newer local project is on-island plastic processing. The local recycling depot has historically hauled all of our plastic waste via large diesel trucks on much larger diesel ferries to processing centers on the mainland; meanwhile other trucks are coming back on those same ferries with new plastic goods. I’m working with them to, where possible, use machinery designs from Precious Plastics to turn plastic waste into sheet goods and from there into useful products that can be used locally like furniture and construction materials. The project is very early (the machines are still being assembled) but I’m excited about it for next year.
Housing is a huge problem on the island, as in many places. Prices have at least doubled since 2019. Rental stock is basically nil (it’s economically more attractive to do short term rentals during the tourist season). Since it’s an island, the local workforce has to be local - it’s impractical to commute from somewhere cheaper. This creates a pretty untenable situation. Building more housing isn’t easy, though: zoning forbids any kind of multi-unit development.
This year I founded the Galiano Cottage Co-op. Our model is to take advantage of the one kind of housing capacity that exists at scale, which is the right to build an 860 sq ft ADU as a secondary dwelling on any multi-acre residential lot (which are the norm). We hope to build a standard cottage design on as many different lots as we can, leasing a portion of the lot for 20 years for a nominal amount in return for fronting the construction cost of the unit; the co-op gets those 20 years of use, after which the cottage reverts to the landowner. We’ll do three builds in 2024, more in 2025, and see where it takes us.
For the latter half of this year I’ve been indulging myself in an expensive passion project: building a self-sufficiently solar powered cruising power catamaran. Maybe it’ll become a business, if enough other people like the design, or provide some useful plans and data to the world that can inspire an even better design. At any rate, I expect it to hit the water sometime late summer of 2024, and to be enjoyably working on it (read: coordinating many others with more skills to work on it) until then. Some specs: it’s based on the composite hulls of a Maine Cat 38 (38’ LWL, 21’ beam), but with a custom aluminum deck house whose roof is 7000W (about 350 sq ft) of bifacial solar panels. It runs on twin electric motors that can push it about 7kts with 8kW of total power. That means you can run at that speed for about 4-5hrs, covering 30 miles or so, with net-zero power use on a summer day. However, it’ll also have about 120kWh of battery allowing longer or faster runs if you take some time at anchor to charge up afterwards.
Finally, I am still doing some software here and there. I’m particularly interested in the relationships between, as a rough list: CAD constraint systems, Bayesian generative models, constructive geometry APIs (including turtle graphics), projector and pen-based UIs, and SDF shaders. I have a lot of thoughts and prototypes around this stuff and want to keep doing more.
If any of this stuff is at all interesting, please feel free to reach out at avi@avibryant.com. If you’re looking for funding for a related project, I can often help find it. Also worth mentioning: I run occasional week-long retreats on Galiano if you’re interested in bringing a group out in person (whether to work on this kind of stuff or anything else, really).