In a slight departure from my usual amateur radio topics, I’ve been busy setting up more smart home stuff. I’ve been curious about home automation for several years and slowly accumulating things like Google Home speakers, Leviton dimmer switches and Nest outdoor cameras. However, using the Google Home app as the central integration point and relying on WiFi for communication for everything has been showing its limitations lately.

Enter Home Assistant, an open-source home automation platform that focuses on local control and privacy. I have several work friends that have rave reviews of Home Assistant, so I decided to give it a try. This has required revamping a lot of my basic home IT infrastructure: I picked up a used workstation from The Server Store to use as a home lab server (for more than just HA), and have replaced my Google WiFi mesh network with a Unifi setup. To add communication options, I also picked up a GoControl HUSBZB-1 USB stick that supports both Zigbee and Z-Wave.

From there, I’ve been setting up Home Assistant on the server, adding integrations for existing devices and services, and adding new devices. Most things have been wonderfully easy. Considering it’s open source, Home Assistant is really well polished in terms of usability and seamless integration. I’ve added quite a few ThirdReality Zigbee devices like motion, door, and temperature sensors, wall plugs, and some Innr color light bulbs. I was even pleasantly surprised to see that my OctoPi 3D printer server popped right in!

I have some overriding goals for my smart home:

1. Value add. Walking into a room turns on the lights, saving effort; the lights turn off when no one is there, saving energy. Notifications let us know when the garage door is left open or when the girls have dance class. The thermostat prioritizes rooms that are actually in use.
2. Spouse-approved. I don’t want to make things more complicated for my family or guests. Everything should be intuitive and reliable.
3. Fail-open operation. I don’t want to have to worry about my home being unusable if a sensor, the internet, WiFi or local HA server are down. Wall switches and the thermostat still need to do their basic job. Everything needs to revert to a dumb appliance in the absence of smart control.

I wonder how hard it would be to add ham radio info to Home Assistant? “There’s DX on 20m…”

Yesterday, RMHAM hosted their first big ARDF foxhunt in a while. I’ve run two foxhunts for high school students before, but this is the first city-scale foxhunt I’ve participated in.

In addition to the traditional tools of the trade, an HT, yagi and offset attenuator, I was using something special: a KrakenSDR. The receiver consists essentially of five RTL-SDR dongles (five sets of R820T2 and RTL2832U chips) tied to a signle clock source, which lets it calculate the direction of arrival (DOA) of signals. Instead of jumping out of the car and spinning around with the yagi, I can just drive around and let the Kraken sample every few seconds. Even if each sample is not quite as accurate, the Kraken can take many more samples and average them to get a more accurate result.

Yesterday morning as the event started, I felt good about my chances of being successful. When the fox started transmitting, I was able to get some pings and start driving generally towards it. I wish I’d had a separeate driver with me so I could focus on the Kraken, but as it was, I sort of noticed the Kraken’s predicted target was jumping around a little bit as the samples came in. After a couple of miles of driving I turned on the turn-by-turn navigation in the Kraken app and started following it. The pings kept getting stronger, and I was getting closer. Eventually I turned into a suburban neighborhood that was a rabbit-warren of criss-crossing roads; if I had been using the yagi I would have had a very hard time. But the Kraken knew where the fox was and the navigation was spot-on.

The fox was located at a house that was only obvious because of the antennas on the roof. I parked, rang the doorbell, and sure enough, this was the place! I was the first one to find the fox. John W0VG and Mark N0OHJ hadn’t even finished getting their coffee yet. I was pretty proud of myself.

Admittedly, this was designed to be an easy foxhunt. The fox was transmitting at 20W from a pretty high location, without any tricky terrain. Robert N0CFM talked about some of the foxhunts he’s run, and one in particular where he was transmitting from a low canyon and the signal was bouncing off a rock formation outside the search boundary. That would be a lot harder to find even with the Kraken. Still, have to start somewhere!

I’m beginning a project with Willem AC0KQ to build an AllMon replacement for the Colorado Connection. The Colorado Connection is a network of AllStarLink linked analog repeaters spanning across the state of Colorado. It’s a great resource for hams in the state, but the AllMon web interface is a bit dated. We’re calling our replacement ColMon, and it will be customized to our needs.

Willem built a working prototype before I joined the project. We’ve worked together a little bit in the past, and I respect him a lot, so I’m excited to be working with him on this project. That said, we come from different generations of programming. Willem’s systems are usually written in C or Perl, using Apache CGI for dynamic web pages and hand-rolled HTML, CSS and JavaScript, and plaintext protocols with tab-separated fields and newline record separators. I’m more comfortable using Angular TypeScript and Golang, with a preference for REST APIs and JSON.

The analogy I’ve landed on is that he’s used to hand-crafting beautiful log cabins, while I’m ready to bring in the house-sized 3D printer. Is either approach better or worse? No, they both have pros and cons, but we’ll both have to get out of our comfort zones and work together to find a middle ground where we can both be productive.

With some luck, our skills together will come up with a beautiful and innovative fusion like the Cornell Ashen Cabin.