AI0009: A Magic Checkbox, a Cotton Undershirt, and 10 Gig at Home - TelePost.ai

This week is about the parts of a deployment nobody puts on the datasheet. A license plate camera that reads perfectly all day and quits the moment infrared kicks in. Six seconds of a launch video that say more about a vendor's strategy than the product being launched. And the credential you need before a spectrum analyzer is allowed past the gate of a refinery.

Three stories, all from the same idea: the radio is rarely the hard part anymore. The hard part is the power budget, the permission slip, and whether anyone at the vendor picks up the phone when you ask for a feature.

Ubiquiti now pushes unknown license plates via API, and a lesson about PoE Plus that only shows up after dark. Ubiquiti's whole-home mesh pack, the MLO backhaul, and the six seconds aimed at managed service providers. What it actually takes to run a wireless site survey inside a refinery, from the TWIC card to badging out.

Everything in this episode, in full:

• Ubiquiti Put MLO Backhaul in a Home Mesh Pack. The Radios Are the Cheap Part Now
• Before You Survey a Refinery, You Need a TWIC Card and a Cotton Undershirt
• I tried the Ubiquiti LPR cameras & they quickly proved how the hardware, and the people, stepped up! All LPR now via API!

Made with TelePost.

Transcript

Copy Transcript Copied

Refinery Survey Safety That Matters

SPEAKER_01 0:30

If you're wearing a nylon or a spandex undershirt inside a refinery and you get caught in an accident, that shirt melts into your skin just from the heat. So the undershirt has to be 100% cotton or flame retardant rated. That's the level of detail nobody warns you about before you walk in with a spectrum analyzer. This podcast is generated by telepost.ai using my words, my voice, and my ideas. Check out telepost and make sure to tell a friend. I'm Drew Lentz, and this is the Waves Weekly Rundown where I talk through what I saw this week in Wireless, in broadband, and in the work itself. Four things today. What it really takes to survey inside an oil and gas facility, why six seconds of a Ubiquity launch video caught my attention more than the product did. What Two Days in Houston told me about an industry that's outgrowing its own vocabulary and the three staffers I was about to hire and never did. Every habit you've built doing ordinary site surveys gets thrown out at the gate. Bring your phone, walk where you like, drive yourself in. None of that applies. What replaces those habits is credentialing PPE and safety protocols, and those are the things that keep you alive. They're also the things that keep you invited back. Start with the Twick card, that's a transportation worker identification card, and enrollment is four steps. So you file online, you book an appointment at a local enrollment center, IdentiGo is the national brand. You sit the background check and give them thumbprints and fingerprints. A couple of days later, the card shows up in the mail. It's the same service as TSA Precheck because it's part of TSA, but the relationship only runs one direction. Getting a twick qualifies you for pre-check. Being pre-checked does not qualify you for a twist seat. What that card buys you is the ability to get on the property unescorted. Without it, somebody from the plant has to walk with you the entire time. Then there's the kit, and it's not negotiable. Steel-toed boots, you know, eye protection, goggles or glasses, and sunglasses usually won't work. Hearing protection, a hard hat, a safety vest. And in oil and gas, depending on the environment, flame retardant clothing, FR2 or FR3 level. You've got two ways to get there. You can dress up like a Ghostbuster, which is what I did, and get a jumpsuit that covers everything. Those coveralls are kind of expensive, or you get flame retardant jeans. Wrangler makes some and pair them with a flame retardant shirt. Some places will inspect you for FR socks, some will inspect you for FR underwear. Ah, check in, they want the TBWick card and the pin that goes with it typed into a keypad, then a six to ten minute training video. No test at the end, at least not that I've seen, but you're expected to be aware of what's going on around you. Here's the one people overlook. Your phone is contraband until somebody says otherwise. Cameras aren't allowed in facilities like this for worker safety, for distraction, and for electrical discharge. A camera phone is still a camera. So if you're taking pictures using distance monitors, carrying any electronic device at all, you need written commitment from the safety officer that you're you're approved to bring it in. Even a spectrum analyzer. Otherwise, they aren't letting your tools through the gate. Once you're out there, do not go off the beaten path ever. There are things there that won't just kill you, they'll make you really, really sick. Yellow bollards, stay inside them. Yellow chains, stay outside them. Don't touch any water or any liquids. If you see people working in those bright silver flame retardant suits, stay away from them. And red and white tape, get away from it as fast as you can. H2S will cause you to stop breathing immediately, and there's no warning for it. That's why you wear the monitor. They'll usually lend you one, and if you don't turn it back in, they'll charge you three to five hundred dollars for it. One last thing that matters more than it sounds. Badge into everything everybody else badges into and badge back out. If they run a security check and see someone badged in but never out, that's a problem for the plant. Because a problem for the plant, and the odds of getting invited back are very small. The thing I've learned doing these is that the protocols aren't there for their insurance. They're there for you.

Ubiquiti Mesh And The MSP Signal

SPEAKER_01 4:17

There are six seconds in Ubiquiti's launch video that I keep pausing on, and they have almost nothing to do with the mesh node. Ubiquity's been around a couple of decades. They've finally entered a space with a product pointed directly and competitively at whole home mesh networking, and it's a doozy. Their gear has always worked well. Over the last couple of years, there have been changes in capability and functionality, some great stretches and some less solid ones. But lately the track record is really good. So them entering the space is a bigger signal about what's coming. Because here's the thing there's never really been an easy button with Ubiquity. The MicroTick companies like that, they make low-cost, feature-rich Wi-Fi devices that still carry a learning curve, and now they're going after a part of the industry that is notoriously user-centric and notoriously customer service heavy. I have no insider knowledge here, but if you've got the confidence to walk into that market at this level, what else is coming? Now, the six seconds, it was their push into the managed service provider space, and it got about six seconds on the reel. Are they going after internet service providers, wireless and fiber providers, college campuses and dormitories? That's what it looked like to me. MSP packages for multi-user installations, all the features they've built for internet service providers at a macro level, all the enterprise and VAR channel work now sitting in the same context as whole home Wi-Fi. What I'm still waiting for is the fabric announcement that congeals all of it. They talked about fabric, they talked about configuration. The missing piece is the thing that ties it together. One canvas for everything ubiquity. When they do that, it's an interesting moment for the industry. But is it actually simple or does it just look simple to us? People who use this gear will tell you it's simple to install. Those are nerds talking, those are people comfortable in CLI. Simple for them doesn't hit the same way as simple for your grandma or your mom or your dad trying to set up a network at the house. And if it doesn't land there, we'll know very, very soon. Then there's the 10 gig backhaul question. Who at home actually needs that? Gamers, content producers, creators, basically Gen Z, and a lot of people doing AI work, massive amounts of connectivity between areas, and a need to know it holds up. Where it gets more interesting is the ecosystem around it. At WLPC, they brought out the wireless switch, eight switch ports backhauling over five and six gigahertz into your existing ubiquity infrastructure. Think about a pool house, a larger property, a ranch, somewhere you need a switch on the far side of the land meshed in, that bridges home and business. And don't underestimate who's already holding this gear. A lot of VARs, the IT person down the street, business owners who already run it at the shop. If a module from managed service providers lets that person turn an extra dime serving the apartment complex down the road or the local university that changes their dynamics completely, they'll push it harder. One practical wrinkle, six gigahertz behaves differently than five. You don't space nodes as far apart, you don't put as many walls between them. Some installers are going to put two nodes up, call it a day, and then discover they need more density. But with MLO bonding into five gigahertz, you get the best of both. Close in, great connectivity, further out or through walls. Another radio picks it up, but I haven't used it yet. I'm more curious about what the public says once these are deployed in every kind of house there is.

Broadband Moves Past Infrastructure

SPEAKER_01 7:31

Two days at the broadband community summit in Houston convince me the conference isn't the one having an identity crisis. The industry is, and not because we don't know what we're doing. Quite the opposite. We've what we've gotten remarkably good at the infrastructure problem we spent decades trying to solve. The questions still open have moved to what happens after somebody's connected. Nobody has answered who pays for that part. We can find money to trench fiber, we can finance infrastructure, we can buy equipment, but who pays for the digital navigator five years later? The show floor felt scattered at first. One side was all MDU. Apartment owners, property managers, managed Wi-Fi, bulk connectivity, resident experience. The other side was municipalities, rural communities, nonprofits, digital equity advocates, funding conversations. By the end, I'd stopped reading that as a programming problem. The center of gravity is moving beyond infrastructure. There is proof the second half can work, by the way. In Far Texas, the city's network hit operating breakeven in about two years at roughly 13 in 100 subscribers. It got there by pairing affordable service with adoption work, digital literacy training, devices. Brandon Dinsmore of NDN Development Group put it simply. Just getting connected doesn't quite solve all the problems. Stephanie Georges of Meraki Dignity Project took it a step further. Access doesn't create the possibility, it's the solutions that leverage the technology that make things possible. Let me be blunt about that. A fiber passing doesn't change someone's life. A Wi-Fi signal doesn't change someone's life. A speed test definitely doesn't. What somebody can do with that connection might. Heather Gate from Connected Nation made a point that stuck with me. We were at a conference called Broadband Communities. We talk about broadband all day. Walk outside and ask 10 random people about their broadband and how many even use that word. Ask about their internet, and they know exactly what you mean. That sounds semantic. I don't think it is. Broadband described the infrastructure challenge that the the people we're serving aren't thinking about infrastructure. They're thinking about homework work, seeing a doctor, starting a business. That's connectivity. And what comes after connectivity is opportunity. Here's what worries me. A second digital divide is forming. Just because someone can use an iPhone doesn't mean they're digitally literate. Just because someone can scroll TikTok doesn't mean they understand cybersecurity or AI or online privacy. The first divide was about who could get connected. The next one is about who knows what to do once they are. And nobody funds the long tail. Who teaches someone to recognize an AI-generated scam? Who helps a senior get to telehealth? Who helps a small business owner understand what AI could do for them? Who stays? Laurel Leverier from USDA gave one of my favorite answers of the whole summit. She described farmers and ranchers as completely technology neutral. They want all of it. Fiber at the house, wireless across the property, satellite where neither reaches. The farmer doesn't care about our technology argument. They care whether it works where they need it. So I kept landing on four legs. Connectivity, because the network has to exist. Devices because people need something useful to connect with. Education because they need the skills to use it safely and productively. And workforce development because all of that has to translate into jobs and businesses. Miss one and the stool gets unstable. Education is the leg we've underestimated most. I had a great conversation with Connected Nation about putting younger people in position to be the technology resource inside their own communities. Picture a community center where a teenager teaches an older resident to spot an AI fake, and then the older resident teaches that kid how to write a thank you note. That's community, and it's a far more interesting outcome than a coverage map turning green. Texas

AI Agents That Stop Hiring

SPEAKER_01 11:07

has 254 counties, different websites, different file types, PDF, HTML, CSV, whatever's there. He said, sure, we can find the funding and get a staffer or two to gather it. And I said, okay, but then we also need somebody to normalize all of that into one format we can actually use. Then I got home. A friend had just told me about GrokBots. So I built one, named it George Washington, and gave it the exact same charge I'd given the guy who was going to find me staff and funding. I said, General, go to all two Fifth Four County websites in the state of Texas, look for this data, bring it back. Then I spawned another one. That one was obviously John Adams. I told it, when General Washington's done, normalize the data sets as he uploads them into this Google Drive. Then I went back to George and gave it the important command. Spawn as many agents as you need. Next thing I know, I've got Benjamin Franklin, James Monroe, Abraham Lincoln, a whole cabinet doing the work. It was fun and it, you know, it understood what it needed to do to satisfy the requirement. So what did that actually save? I'd avoided hiring three staffers, sure, but the real cost isn't three salaries, it's insurance, payroll, taxes. Honestly, though, it wasn't about the money. It was the time. I didn't have to train anyone. I didn't have to show them what to do, watch them get the data wrong and go back and fix it. Under 24 hours from start to having what I needed back on the core project, writing the job description alone would have taken 24 hours, never mind finding the people. Here's how it felt. I drive a Ford Lightning the other day, the right lane was closing, and the guy on the left kept speeding up to match me every time I did. So I tapped the accelerator, popped out in front of him, gone. Easy, like nothing, because he was in a gas-powered vehicle and I had that power sitting right there. That's what these tools do for me every day. But only because I'm immersed in them, I don't work without AI as my sidekick anymore. And this is where I have to admit I got something wrong. I believed the promise that AI would level the playing field because it makes everything easier for everyone. What I didn't account for was that the people who were already really good would get so much better. The Texas Project is my evidence. The difference wasn't budget, it wasn't headcount, it was knowing what to reach for. So when people ask whether AI is going to take jobs, I think that's the wrong frame. It doesn't take the job, it makes the job never exist in the first place. There's nothing to take away because it was never posted. And that's what puts you on one side of the fence or the other. What do you do about it? Learn the tools that exist right now, not the ones you read about last year. Find a local friend group, people who are excited about this, one or two friends using it in ways you hadn't considered. Get really good at at least one tool until it's your superpower. And the next time you're about to write a job description for multi-step data gathering, try spawning agents first.

The Real Scarcity And Final Takeaways

SPEAKER_01 13:42

Four different rooms this week a refinery, a launch video, a conference hall in Houston, and my kitchen table. And the same thing kept showing up in all of them. The hardware is the easy part now. The radios are cheap, the fiber's getting built, the agents will do the gathering. What's scarce is knowing what to reach for and who's going to stick around and teach the next person. And if any of this is worth more than the version I just gave you, go read the pieces in full. They've all got the detail I couldn't fit here. I'm Drew Lentz. Stay safe out there. Badge out, and I'll see you next week.