Studio711.com – Ben Martens

Message Privacy

These days, companies are using any data they can get their hands on to build profiles of us, market to us, and sell information they’ve derived about us. It’s mostly unavoidable, but it’s nice to exert some control when possible, which is one of the reasons why I use Signal for messaging. This isn’t a sponsored post or anything, but I thought I’d share in case you didn’t know about this but would find it helpful.

Text messages are unencrypted. Anybody in between you and the recipient can read them. I’m not saying telephone companies are mining data out of text messages, but I assume they are regardless of what their privacy policies claim. At the very least they have the ability to do it. Text messages are where I’m the quickest to say something that could easily be taken out of context and even if nobody is reading them now, I’d rather not have them sit around for years.

The three main reasons that I now use the Signal messaging app are:

  • The app encrypts your data end to end. Even the Signal employees can’t read your messages.
  • Each message thread can be configured to delete messages after a certain amount of time. I usually set mine to 4 weeks
  • Pictures and videos are sent at much higher quality. No more being annoyed at low quality images when sharing between iPhone and Android users!

Apps are available for most platforms. I use it on my phone, but I also run the Windows app on my desktop. There’s no cost to use it. The group that runs it is a non-profit that exists on funding from investors and donations from users.

The only downside to using it is that the people you chat with need to be on it for it to be useful. Thankfully, that’s the case for most of the people I chat with, but some chats still happen in an SMS app. It’s not that big of a deal to just tap the notification for whichever app received the message and open it up.

So chat however you want, but if you get a little itchy when you think about how companies are potentially using your data, this is a really easy way to decrease your attack surface.

Indiana Vacation

This summer we made another trip back to my parents’ house in Indiana. We spent a lot of time in the pool, playing disc golf, and eating delicious food. We also visited Indiana Dunes National Park and saw my grandpa. I put together a quick video recap of the trip and while it’s not going to win any awards, it’s a fun way to look back on the trip.

Tesla Range

Welcome to another Tesla Tuesday!

We recently put 2500 miles on our Tesla Model Y and the road trip was a great experience. However, there was one stretch where I had some range anxiety for the first time. We left Twin Falls, ID with 82% charge which should have been enough to get us into Meridian, ID with 25% remaining. However, as we set out, our range was decreasing rapidly, and I was predicting that we’d get there with around 10-15% remaining. That’s still plenty of buffer, but it was disconcerting. I say this not to scare anyone off of EV’s but to make sure I’m not all roses and butterflies when I talk about driving one. I haven’t been able to figure out exactly what happened, but I suspect a combination of the following based on the minute-by-minute data that was logged from the car.

  • It was HOT. Temperatures were between 99 and 104 on that stretch of the trip. HVAC can consume quite a bit of battery when it’s that hot.
  • The speed limit was 80mph and I had the cruise set at 85mph. More on speed’s impact on efficiency later.
  • We were driving into a headwind.
  • Traffic was heavy-ish so there was a lot of slowing down and speeding up.

I expect the car’s estimation to take all these things into account, but something was definitely off. About 5 miles out from the supercharger in Boise, we spotted a Tesla with lights flashing going about 60mph trying to conserve energy and I assume they hit the same issue but didn’t notice it soon enough. My guess is that our trips were flagged by the engineers at Tesla and hopefully our experience will help improve their estimation models.

For my part, I decided early on to back off my speed and set the cruise at 78mph. This felt about as slow as I could go without being an annoying rock in the stream. That quickly helped to stabilize the estimate. Click on the link below to see all the charts from this stretch of the trip.

We made it with no trouble, but it was the first time that I had thought about what would happen if we didn’t make it. Some stretches of these roads also had signs about “no gas for X miles” but it’s easier to bring a gas can to your car than it is to bring your car to a charger.

I do think it’s interesting how much easier it is to calculate the exact impact of driving at different speeds and heats. When we cruise at 85mph, we were using 360-380 watts per mile but when we cruise at 65 it’s more like 250-270 watts per mile. That’s a ~40% increase in energy usage for an extra 20mph. This same thing happens in a gas-powered car too, but with a bigger tank and less data, you don’t always notice it as easily.

While TeslaMate does show me average efficiency by temperature, it doesn’t have a dashboard for efficiency by speed. It’s tricky to calculate because of all the various factors at play (temperature, elevation, traffic impact, driving style, etc), but if we take a naive approach and look at average efficiency in kWh based on average speed in mph for an entire trip segment, it looks like the chart below. Take this with a grain of salt because as I mentioned, this is a very rough estimate. I would expect an exponentially increasing line. Below about 35mph, there’s not a lot of data since I only used trips of at least 20 miles and I don’t often drive that far that slowly.

With a gas car, your efficiency peaks around 45mph and at slow speeds, you waste a lot of extra energy, but with an electric car, you can go almost as slow as you want and keep using less energy. However, at high speeds, they act similarly: drag increases with the square of speed. Driving a car with a very low coefficient of drag does help, you’re still bound by the laws of physics.

CNC Cherry Bowl

In 2016, we decided to take down two scraggly cherry trees along our driveway. Check out 3-year-old Elijah looking on as Logan and I took the trees down.

There was hardly much wood in them, but I saved some of the bigger pieces anyway. After drying out, one piece of the wood got turned into a cross that hangs in our front entryway. But I had another half dozen or so pieces sitting around with no plan. I had been thinking about trying to make a bowl on the CNC anyway, so I decided to use this wood for that project.

I spent a lot of time with a segmented bowl calculator trying to figure out the specific angles I’d need to cut 8 trapezoids for each layer. Eventually I felt good about the calculations and since that program spits out SVG files, I decided to just cut them on the CNC and know that the angles would be perfect instead of trying to do it on the table saw. The table saw would have been a lot faster, but one bonus of the CNC was that I could do a facing operations on each board to get them to exactly the same thickness. Using the planer for that would have been difficult since the pieces were too short for the planer.

I glued each layer individually and clamped it with hose clamps. Once each layer was done, I glued them all together. Then it was time to model the actual bowl in Fusion 360. The trick was that I had to make sure that the bowl perfectly fit inside the blank. Here again, the SVG files from the calculator came in handy because I could pull them into Fusion and draw the blank pretty much exactly as it existed in the real world. I also had to be careful not to make the bowl too steep since my router body would crash into it while carving out the bottom of the bowl.

Then it was time to program the actual cut in Fusion. I started with the bowl upside down. I programmed a dummy cut around my stock which would give me confidence that I had things calculated correctly. Then I did a roughing pass with a 1/4″ end mill to get rid of the bulk of the material. The picture below shows that roughing pass about halfway complete. The whole pass took 1.5-2 hours.

I then switched to a 1/4″ ball nose bit and took very small passes back and forth across the bowl to give a pretty smooth finish. It still needed sanding, but it was very close to the final surface.

Next I flipped the bowl over and did the same process on the inside of the bowl.

While doing the inside of the bowl, I had an accident where the bit plunged way down into the blank where it shouldn’t have. The origin of my model had been set incorrectly. Chatting about it with Luke, he suggested that I fill the hole with a dowel, and I happened to have a cherry dowel that was exactly the right size. Thankfully that whole mistake ended up getting cut away, so it didn’t impact the final product.

It was also a challenge to hold the bowl to the surface of the table. Since I was machining the whole surface, there was nothing to clamp to. So instead, I put blue painters tape on the bottom of the blank and on the table. Then I glued them together with super glue. It created a strong bond that I was able to break later without damaging the bowl or the CNC table.

I added some boiled linseed oil as the finish and now it’s done! I don’t have any big plans for the bowl, but I’m happy to have made it successfully through this project. I didn’t really have enough wood to make another one if I screwed up and I was too lazy to make a duplicate blank out of cheaper boards.

If I ever do this again, there will be a few changes:

  • I didn’t rotate the second layer correctly. The cracks line up with the base/first layer. It doesn’t look too bad from the inside of the bowl, but it looks goofier from the outside.
  • I should have taken more time sanding the layers flat before gluing them together into the blank. I thought I had them flat enough, but there are some segments that have a tiny gap to the segment above them.
  • I need to experiment more with the program and how fast I run the bit. This whole bowl was something like 6 or 7 hours of machine time plus all the setup time around it.

Surviving Election Season

I’m less and less enamored with politics as time goes on, but it seems unavoidable as we head into presidential election season. Since I know this is a common frustrating for people, I thought I’d share some of the things I do to try and stay sane:

  • Beware of people who use name calling. They’re trying to get you on their “team” and create division. Even if you happen to agree that the other person deserves it, it’s not healthy.
  • Beware of people who try to play to your emotions to get you to agree with them. Reading a news story or watching a talk show shouldn’t be like getting riled up about sports.
  • Beware of people who speak in absolutes or hyperboles. AI isn’t going to kill us all. Their opponent getting elected won’t be the end of the world.
  • Remember that even the choice of news sources can be full of bias. Not reporting on a story can be just as bad as taking a slanted view of the topic.
  • Think about how concerned you are about telling people what you think about a politician. Imagine if you felt that strongly about sharing Jesus with them. What if for every conversation you had where you shared your political opinions, you also had one where you talked about Jesus? Which conversation is more relevant and important?
  • In a world where you want to tell everyone about Jesus, why would you alienate more than half of everyone you meet by telling them about your political opinions or even hinting at them?

Estimating Road Trips

Welcome to anotherĀ Tesla Tuesday!

My big post about our long road trip covered a lot of data, but later, I wondered about our average speed including all the stops. When I planned long trips with the family, I would shoot for 60mph including stops. On my motorcycle, I planned 50mph because I would stop a lot more often. Even with some silly stops where we knew we were being very inefficient just to take things slower, our 2000+ mile road trip averaged out to 58.1mph. So for our family, driving an EV gets us there just as fast as driving a gas car. Your mileage may vary depending on the size and strength of your bladders!

EV Charging Network

Welcome to another Tesla Tuesday!

Over the last few weeks, a flood of car makers and charging networks have stated that they will adopt the NACS charging connector, otherwise known as the Tesla connector. Previously there were two main standards for fast charging: NACS and CCS. There were also a flood of slower chargers.

Having all the automakers in North America adopt the NACS connector means that Tesla will proceed with opening their supercharging network to all electric vehicles as they’ve done in other parts of the globe. That’s great for non-Tesla owners as the Tesla supercharger network is in a different league compared to other networks. It’s extremely rare to pull up to any Tesla charger and find it working. Conversely, it’s rare to pull up to something like Electrify America and find all the chargers working. We just completed a 2500-mile road trip in our Model Y, and while I was confident in the Tesla network, I don’t know that I would have even attempted it with a non-Tesla purely because of the lack of reliability in non-Tesla charging networks.

Having all the networks use the same plug is great, but it doesn’t mean that the other networks are going to be any more reliable. That needs to be a major focus for them in the coming years. The infrastructure needs to be rock solid for people to trust it. They need to not only adopt the charging standard, but the reliability standard that Tesla has set.

As a Tesla owner, I want to see the industry grow and build standards, but it’s a bit of a bummer to have “muggles” at my supercharger. Part of the reason I paid a premium for a Tesla is specifically because of the fantastic charging network. Now it’s more likely that I’ll pull up to a station with no empty chargers. That should be a temporary problem, and with a single standard, we should start to see more people getting creative with their charging stations. I look forward to the day when I can pull into a standard travel plaza on my road trip and charge my car while I use the restroom, get some food, and clean my windshield.

JBLM Airshow and Warrior Expo

As a child, I remember going to some big airshows, and I still have a poster on the wall in this room of the Thunderbirds. Dad stood in line with me to get it signed by all six pilots. So when I heard that Joint Base Lewis McChord (JBLM) was bringing back its airshow after a seven year break, I really wanted to go.

We booked a hotel the night before even though it’s only 60 miles away with the thought of making our morning a little less hectic. Plus, Tyla and Elijah love hotel pools and hotel breakfasts. Elijah thinks Best Western is the most amazing hotel chain ever, and I’m not in any hurry to dissuade him! We did have a relaxing evening on Friday and then Saturday morning, we drove about 5 minutes across the interstate and onto the base. We were directed to park in a field and hopped on a shuttle quickly after that.

The airshow area was huge and there were so many static displays that we couldn’t see them all in the ~2.5 hours before the show started.

The show started off with a “joint force demonstration” since it’s a joint army and air force base. The list of planes that flew included B-17 Flying Fortress, B-25, C-17, Focke-Wulf 190, P-51 Hope, Tora Tora Tora Mitsubishi A6M “Zero”, Undaunted Air Act (Vans RV-7 and RV-8’s), and Yellow Thunder (twin AT-6 Texans). The army also had Apache, Blackhawk, and Chinook helicopters along with a Striker vehicle. I’m sure I’m missing some more!

It was the first time I remember seeing the Tora Tora Tora act. They reenacted the attack on Pearl Harbor with eight planes. Even with only a few planes, it felt like chaos because they had lots of big pyrotechnic explosions happening on the ground while the announcer talked about the attack. It’s impossible to imagine what it would have felt like with 20 times that many planes attack constantly for TWO HOURS.

Of course, the highlight of the whole show was the Thunderbirds! All the acts were fun to watch but the Thunderbirds are the pinnacle of air acts. It was beautiful to watch them with Mt. Rainier standing proudly in the background, and I was so happy to share the experience with Tyla and Elijah!

I was also really happy that we remembered to bring ear protection. It’s LOUD and sitting out there all day in the hot sun is a recipe for a headache even without the loud planes.

While I don’t know if I would have enjoyed lugging my dSLR around all day, I do wish I had it for taking pictures of the airshow. We had a beautiful view with Rainier in the background and while there are plenty of photos online, it’s always fun to snap “the shot” yourself.

We had gone extra fancy for this event and paid for reserved seating. It was really nice to have “saved seats” and not have to fight for position. We also had easy access to bathrooms and food. I’d make that same choice again.

I’ve been wanting to go to an airshow for years and I’m so thankful that it finally happened!

Ok so that’s the happy side of the day. Let’s talk about the downsides.

  • It was HOT. Temps were up around 90 and since you’re on an airfield, there’s no shade. We brought water bottles and refilled them multiple times at bottle filling stations that they had cleverly hooked up to fire hydrants.
  • Traffic was bonkers. The news said that the airshow hit capacity around 1:30pm (about halfway through the air acts). Thankfully we missed it going in but it bit us leaving.
  • We had to take a shuttle back since our car was ~8 miles away. I made the tough decision to leave the show about 2/3 of the way through the Thunderbirds act because I knew the shuttles were going to be a mess. While I hated leaving early, we were able to still see most of the maneuvers as we walked back. We arrived to CHAOS. It was unclear where the busses were going to stop and which line was for which shuttle. The lines were already enormous, but thankfully the line for our shuttle was shorter. This was part of my plan by picking the shuttle on the south end because I knew the north end would be swamped with all the people from Seattle. We waited baking in the hot sun in a line of angry people for an hour. That got us on the THIRD BUS. Yes, the busses were incredibly slow because they were getting caught up in all the traffic leaving the show. I’d estimate that we were about 5-10% of the way from the front of the line and it took us an hour to leave. People must have been there for hours. Thankfully, there were quite a few military personnel there to help sort out line cutters and other squabbles.
  • They did have limited parking on base so I suppose we could have tried that, but then you’re resigning yourself to sitting in traffic for hours trying to squeeze out of the one lane gate off the base.

I honestly don’t know what I’d do if I could do this again. I think the winning move was probably parking at the closest shuttle stop (SR-512 Park and Ride) and then just walking the ~3 miles back to that shuttle stop after the show. But then again, the line for that shuttle stop was enormous in the morning because the city buses went to the wrong location to start the day!

Both our bodies and nerves were fried by the end of the day, but I was so proud of Elijah for sticking with it all day! If anyone in our family had given into a bad attitude, it would have been miserable for all of us, but we all left thankful that we saw what we saw, but also very thankful to be sitting in air conditioning and driving home. I’m not eager to go again, but I’m glad we went!

GPT and LLMs

ChatGPT made a big splash last fall and large language models (LLMs) in general continue to be a hot topic. My current favorite is using Bing AI (https://bing.com/chat) because it combines current search results with the information it already has in its model.

But there’s still a lot of confusion about what is powering these sites. How do they know things? Are they sentient? Stuff You Know Should Know did a great episode called Large Language Models and You and I think it’s worth a listen. I’ve heard LLMs previously described as “word salad”, and this episode gives another good example of explaining that it’s like an iteration on autocomplete in a text box. The algorithm just knows what words are most likely to come next and which words are related to each other. It has no concept of what it is saying. It only knows that those words are most likely to go together when it sees your prompt. So there’s no sentience or actual knowledge happening here, which is probably good but it’s also bad because it means that ridiculous answers can come out and be presented as fact.

The episode covers all of this and then also does a good job of how incredibly fast things are improving. Give it a listen if this topic interests (or scares) you.

1100 Miles in a Tesla

Welcome to another Tesla Tuesday!

Up until our trip to Moab, we had never done a trip that required multiple charges away from home, but we sure put that behind us with this trip. Our total drive was around 2500 miles there and back. The video at the end of this post covers the drive to Moab.

There are a few things that I didn’t cover in the video:

  • Supercharging is almost irrelevant for daily life. I wrote this whole post and made a video about this because we’ve gone 20,000 miles in our car, and we’ve only used a supercharger a few times. Charging like this is a novel experience for us! I often hear people say, “Oh I can’t get an electric car because there are no chargers around me.” The chargers near you don’t matter. You’ll probably never use a charger within 100 miles of your house because you constantly have a charged car when you wake up (assuming you can charge where you park.) The only time you need a good charger network is if you decide to take your EV on a long road trip, and even then, if you have two cars, you could just drive your gas car. So the “what’s it like to charge an EV?” question is something I thought a lot about before I got the car, and now I hardly think about it at all. My car is always charged and it requires about 5 seconds to plug it in when I get home. It only comes into play when we’re taking a long road trip.
  • Charing at a supercharger is extremely easy. When we stopped to charge, the car was charging before my family even got out of the car. I’d park, hop out, grab the charger, and plug it in. There’s no fumbling with a credit card or an app. The charger communicates with the car and automatically charges the credit card associated with the car. The car knows how far it needs to charge before you’re ready for the next leg of your trip or you can manually set your own limit.
  • Very little of the planning that I show in the video is something a normal Tesla driver would do. You could easily complete this trip while being completely oblivious of all the optimizations that I made. When you punch in your destination, the car tells you where to stop and how long to charge at each stop. It will give you many warnings if you try to leave the charger too soon or if something changes en-route and you’re unlikely to make it. This short video from Tesla explains how that works:

I had a data logger recording every datapoint from our car every few seconds and I took some video during each leg of the trip. I put it all together into a video about the trip along with my thoughts about whether I’d do it the same way again. Enjoy!