Wong Edan's
Cover
IoT & Hardware

IoT, Industrial Grunts, and Why Maketronics is Finally Breathing

March 21, 2026 • By Azzar Budiyanto

The Madman Returns: IoT, Industrial Gremlins, and the Maketronics Genesis

Listen up, you keyboard-mashing code-monkeys and solder-sniffing hardware junkies. It is been a busy year—actually, let’s be honest, it has been a chaotic dumpster fire of productivity. If you have been refreshing the Flashgamer feed wondering if I’d fallen into a vat of molten lead-free solder, I have news for you: I am still alive, and I have been building things. Real things. Things that actually do stuff besides blinking a single LED and making me feel like a wizard for five minutes before the magic smoke escapes.

The big headline? Maketronics has finally birthed its first real product. On August 16, 2024, the world (or at least the three people paying attention) witnessed the official announcement. It is been a long time coming. I have spent months neck-deep in IoT, industrial sensors, and the kind of technical debt that makes a bank’s balance sheet look like a toddler’s coloring book. This article is a deep dive into the madness—from the RadSense 1 launch to why your USB-C cables are probably lying to you.

“It’s been a busy year, so I’ve not written much about ongoing projects…”

That is the understatement of the century. But grab your multimeter and a stiff drink, because we are going under the hood of the Maketronics ecosystem and the industrial IoT landscape that is currently trying to melt my brain.

1. The RadSense 1 Launch: Lessons Learned in the Hardware Trenches

Maketronics didn’t just appear out of thin air. It was forged in the fires of “Why doesn’t this work?” and “Who designed this footprint?” The RadSense 1 launch was the culmination of everything we have been trying to achieve with industrial-grade sensing in an IoT package. Launched officially in mid-August 2024, the RadSense 1 represents a shift from “project” to “product.”

What did we learn? First, hardware is hard. You can simulate everything in KiCad until your eyes bleed, but the real world is a cruel mistress. The RadSense 1 taught us about the logistics of sourcing components that aren’t counterfeit and the joy of designing for manufacturing (DFM). We had to develop new tools just to keep up with the testing requirements. When you are moving from a breadboard to a production-ready PCB, the “it works on my machine” excuse doesn’t fly when a customer’s industrial PLC (Programmable Logic Controller) is screaming at you.

The RadSense 1 isn’t just a sensor; it is a statement. It’s about taking those flimsy “hobbyist” sensors and giving them the backbone needed for actual industrial applications. We’re talking about reliability, noise immunity, and a form factor that doesn’t look like it was hacked together by a caffeinated squirrel.

2. The Industrial Reality: ADS Errors and PLC Nightmares

If you think debugging a React app is frustrating, try talking to a PLC. In the industrial world, things are… different. Take the infamous Ads-Error 0x701: “Service is not supported by server.” This error is the bane of anyone trying to bridge the gap between custom hardware and established industrial systems.

When you are building hardware to interface with a PLC, you aren’t just sending bits; you are negotiating a peace treaty between two different eras of technology. To reliably talk to the PLC, you first have to enable specific services on the server side. It is industrial-grade gatekeeping at its finest. My experience with these systems led to a realization: most “IoT” solutions are too fragile for the factory floor. They can’t handle the electrical noise, the protocol strictness, or the simple fact that in an industrial setting, if a sensor fails, it doesn’t just mean a missed notification—it means a production line stops.

This is where the Maketronics philosophy comes in. We’re building the hardware I needed but couldn’t find. Hardware that understands the 0x701 error and doesn’t just roll over and die when the PLC asks for a service it doesn’t recognize.

3. ESP32 and the Holy Grail of OTA Updates

On August 14, 2024, just days before the Maketronics product launch, I went deep into Implementing Over The Air (OTA) updates with ESP32. Why? Because if you have to climb a ladder or open a sealed industrial enclosure every time you find a bug in your C++ code, you have failed as an engineer.

The ESP32 is a beast, but its OTA implementation is where the real power lies. For the Maketronics line, OTA isn’t a luxury; it’s a requirement. We use a dual-partition scheme where the new firmware is downloaded to a “passive” partition while the “active” partition continues to run the industrial processes. Only once the checksum is verified does the bootloader swap them. This prevents “bricking” the device—a situation where the hardware becomes a paperweight because of a corrupted 1 and 0.


// Simplified OTA Check logic
if (Update.begin(UPDATE_SIZE_UNKNOWN)) {
  uint32_t written = Update.writeStream(payload);
  if (written == contentLength) {
    if (Update.end()) {
      if (Update.isFinished()) {
        ESP.restart();
      }
    }
  }
}

The complexity of implementing this in a production environment, especially with IoT security in mind, is what separates the pros from the “I just followed a YouTube tutorial” crowd. You have to handle rollbacks, versioning, and secure handshakes. If you don’t, some kid in a basement will be “OTA-ing” your industrial sensors to play Doom.

4. The Soil Moisture Saga: Capacitive vs. The World

Let’s talk about Capacitive soil moisture sensors. Back in late 2019, I started testing these, specifically the ones found in HiGrow boards. Most cheap soil sensors are resistive—they stick two metal probes in the dirt and measure the resistance. Spoiler alert: they corrode faster than a cheap car in a salt mine. The metal literally dissolves into the soil because of electrolysis.

The sensors we are integrating into the Maketronics ecosystem are capacitive. They use an analog interface and support a 3-5V operating voltage. Instead of passing current through the soil, they treat the soil as a dielectric in a capacitor. As the moisture content changes, the capacitance changes, which we then convert into a voltage signal. This means no corrosion and a much longer lifespan in harsh environments.

However, it isn’t all sunshine and rainbows. These sensors are analog, and ESP32 ADCs (Analog-to-Digital Converters) are notoriously non-linear and noisy. You can’t just plug it in and expect a perfect percentage. You need calibration, hardware filtering (RC circuits), and software averaging to get a reading that actually means something. It’s another example of where “simple” IoT projects meet the brick wall of “industrial” reality.

5. The USB-C Conundrum: Why Two Type C Cables Won’t Provide Power

You’d think in 2024, we’d have figured out how to move electricity through a wire. But no, USB Type-C is a disaster zone. I’ve written about this because it’s a constant thorn in the side of hardware development. You have a device with a USB-C port. You have a high-end PD (Power Delivery) charger. You connect them with a C-to-C cable. Nothing happens. You switch to an A-to-C cable? It works perfectly. Why?

The answer lies in the CC (Configuration Channel) pins. In a proper USB-C implementation, the “sink” (your device) needs to have 5.1k ohm pull-down resistors on the CC1 and CC2 pins. Without these, a smart C-to-C charger looks at the cable and sees… nothing. It assumes nothing is connected and refuses to send any power to protect against shorts. Cheap “A-to-C” cables don’t care because the “A” side is always “dumb” and just pumps out 5V. If you’re designing IoT hardware—like our new microcontroller reference boards—and you forget those two tiny resistors, your “modern” device is effectively dead on arrival for anyone using a modern charger.

6. Missed Opportunities and the LEGO Ghost

Not every project makes it to the Maketronics finish line. I’ve been open about my “missed opportunity” with the LEGO light controller. It’s a classic case of over-engineering leading to paralysis. I wanted a perfect, networked, synchronized lighting system for my LEGO sets. I spent so much time worrying about the protocol and the dimming curves that I never actually finished it.

This is a vital lesson for any tech builder: Perfect is the enemy of Shipped. The reason Maketronics finally has a product in August 2024 is that I stopped trying to make it do everything and focused on making it do its core job reliably. The LEGO controller sits on a shelf as a reminder that a simple, finished project is worth more than a revolutionary, unfinished one. It also informs how we approach our new microcontroller reference boards—focus on the essentials, get the power delivery right, and make sure the GPIOs are accessible.

7. New Microcontroller Reference Boards: The Future Foundation

Speaking of reference boards, we’ve launched a new one to coincide with the Maketronics push. Why another board? Because the market is flooded with boards that are either too basic or too specialized. Our new reference board is built for developers who are tired of fighting with bad pinouts and lack of protection circuitry.

It includes features we’ve learned are essential through our industrial sensor work:

  • Robust power regulation that can actually handle the spikes of an ESP32 WiFi radio.
  • Properly labeled pins that match the software definitions (you’d be surprised how rare this is).
  • Integrated protection for the aforementioned USB-C CC pins.
  • Pre-configured pads for capacitive sensing inputs.

This board is the “internal” tool we used to develop the RadSense 1, and now it’s part of the broader Maketronics ecosystem. It’s about creating a standardized platform so that every time we want to build a new industrial sensor, we aren’t reinventing the wheel—or the power supply.

Wong Edan’s Verdict

So, what’s the bottom line? Is the IoT and industrial sensor world a beautiful, synchronized dance of data? No. It’s a chaotic brawl between legacy PLC protocols, picky USB-C chargers, and moisture sensors that want to commit suicide by corrosion.

Maketronics’ first product, the RadSense 1, is my attempt to bring a little bit of sanity to the madness. We’ve taken the lessons from failed LEGO controllers, the frustration of 0x701 ADS errors, and the technical necessity of ESP32 OTA updates, and crammed them into actual hardware.

My verdict? Hardware is a special kind of madness. It’s expensive, it’s frustrating, and it takes way longer than software. But when you finally see that data from a capacitive sensor hitting your dashboard via a secure OTA-updated ESP32, without a single 0x701 error in sight… well, it’s enough to make even a Wong Edan like me smile. Stay tuned. We’re just getting started, and the lab is still smelling like flux. It’s a good smell. It’s the smell of things actually getting done.