machine learning

241 Articles

Illustrated Kristina with an IBM Model M keyboard floating between her hands.

Keebin’ With Kristina: The One With The (Mc)Cool Typewriter

September 16, 2024 by 2 Comments

A hand and wrist with a gesture detection ring and a control box on the wrist.
Image by [ambrush] via Hackaday.IO
Okay, so this isn’t a traditional keyboard, but you can probably figure out why the RuneRing is here. Because it’s awesome! Now, let me give you the finer points.

Hugely inspired by both ErgO and Somatic, RuneRing is a machine learning-equipped wearable mouse-keyboard that has a configurable, onboard ML database that can be set up to detect any gesture.

Inside the ring is a BMI160 6-axis IMU that sends gesture data to the Seeed Studio nRF52840 mounted on the wrist. Everything is powered with an 80mAh Li-Po lifted from a broken pair of earbuds.

Instead of using a classifier neural network, RuneRing converts IMU data to points in 24-dimensional space. Detecting shapes is done with a statistical check. The result is a fast and highly versatile system that can detect a new shape with as few as five samples.

Continue reading “Keebin’ With Kristina: The One With The (Mc)Cool Typewriter” →

Supercon 2023: Teaching Robots How To Learn

September 3, 2024 by 2 Comments

Once upon a time, machine learning was an arcane field, the preserve of a precious few researchers holed up in grand academic institutions. Progress was slow, and hard won. Today, however, just about anyone with a computer can dive into these topics and develop their own machine learning systems.

Shawn Hymel has been doing just that, in his work in developer relations and as a broader electronics educator. His current interest is reinforcement learning on a tiny scale. He came down to the 2023 Hackaday Supercon to tell us all about his work.

Continue reading “Supercon 2023: Teaching Robots How To Learn” →

Building AI Models To Diagnose HVAC Issues

August 8, 2024 by 10 Comments

HVAC – heating, ventilation, and air conditioning – can account for a huge amount of energy usage of a building, whether it’s residential or industrial. Often it’s the majority energy consumer, especially in places with extreme climates or for things like data centers where cooling is a large design consideration. When problems arise with these complex systems, they can go undiagnosed for a time and additionally be difficult to fix, leading to even more energy losses until repairs are complete. With the growing availability of platforms that can run capable artificial intelligences, [kutluhan_aktar] is working towards a system that can automatically diagnose potential issues and help humans get a handle on repairs faster.

The prototype system is designed for hydronic (water-based) systems and uses two separate artificial intelligences, one to analyze thermal imagery of the system and look for problems like leaks, hot spots, or blockages, and the other to listen for anomalous sounds especially relating to the behavior of cooling fans. For the first, a CNC-like machine was built to move a thermal camera around a custom-built model HVAC system and report its images back to a central system where they can be analyzed for anomalies. The second system which analyses audio runs its artificial intelligence on a XIAO ESP32C6 and listens to the cooling fans running in the model.

One problem that had to be tackled before any of this could be completed was actually building an open-source dataset to train the AI on. That’s part of the reason for the HVAC model in this project; being able to create problems to train the computer to detect before rolling it out to a larger system. The project’s code and training models can be found on its GitHub page. It seems to be a fairly robust solution to this problem, though, and we’ll be looking forward to future versions running on larger systems. Not everyone has a hydronic HVAC system, though. As heat pumps become more and more popular and capable, you’ll need systems to control those as well.

AI Kayak Controller Lets The Paddle Show The Way

June 10, 2024 by 6 Comments

Controlling an e-bike is pretty straightforward. If you want to just let it rip, it’s a no-brainer — or rather, a one-thumber, as a thumb throttle is the way to go. Or, if you’re still looking for a bit of the experience of riding a bike, sensing when the pedals are turning and giving the rider a boost with the motor is a good option.

But what if your e-conveyance is more of the aquatic variety? That’s an interface design problem of a different color, as [Braden Sunwold] has discovered with his DIY e-kayak. We’ve detailed his work on this already, but for a short recap, his goal is to create an electric assist for his inflatable kayak, to give you a boost when you need it without taking away from the experience of kayaking. To that end, he used the motor and propeller from a hydrofoil to provide the needed thrust, while puzzling through the problem of building an unobtrusive yet flexible controller for the motor.

His answer is to mount an inertial measurement unit (IMU) in a waterproof container that can clamp to the kayak paddle. The controller is battery-powered and uses an nRF link to talk to a Raspberry Pi in the kayak’s waterproof electronics box. The sensor also has an LED ring light to provide feedback to the pilot. The controller is set up to support both a manual mode, which just turns on the motor and turns the kayak into a (low) power boat, and an automatic mode, which detects when the pilot is paddling and provides a little thrust in the desired direction of travel.

The video below shows the non-trivial amount of effort [Braden] and his project partner [Jordan] put into making the waterproof enclosure for the controller. The clamp is particularly interesting, especially since it has to keep the sensor properly oriented on the paddle. [Braden] is working on a machine-learning method to analyze paddle motions to discern what the pilot is doing and where the kayak goes. Once he has that model built, it should be time to hit the water and see what this thing can do. We’re eager to see the results.
Continue reading “AI Kayak Controller Lets The Paddle Show The Way” →

An Improved Spectrometer, No Lasers Required

May 27, 2024 by 23 Comments

Here at Hackaday, we love it when someone picks up the ball from a previous project and runs with it. That’s what we’re all about, really — putting out cool projects that just might stimulate someone else to extend and enhance it, or even head off in an entirely new direction. That’s how the state of the art keeps moving.

This DIY spectrometer project is a fantastic example of that ethos. It comes to us from [Michael Prasthofer], who was inspired by [Les Wright]’s PySpectrometer, a simple device cobbled together from a pocket spectroscope and a PiCam. As we noted at the time, [Les] put a lot of the complexity of his instrument in the software, but that doesn’t mean there wasn’t room for improvement.

[Michael]’s goals were to make his spectrometer a little easier to build, and to improve the calibration process and overall accuracy. To help with the former, he went with software correction of the color filter array on his Fuji X-T2. This has the advantage of not requiring a high-power laser and precision micropositioner to ablate the CFA, and avoids potentially destroying an expensive camera. For the latter, [Michael] delved deep into the theory behind spectroscopy and camera optics to develop a process for correlating the intensity of light along the spectrum with the specific wavelength at that location. He also worked a little machine learning into the process, training a network to optimize the response functions.

The result is pretty accurate spectra with no lasers required for calibration. The video below goes into a lot of detail and ends up being a good introduction to some of the basics of spectroscopy, along with the not-so-basics.

Continue reading “An Improved Spectrometer, No Lasers Required” →

Tabletop Handybot Is Handy, And Powered By AI

May 21, 2024 by 15 Comments

Decently useful AI has been around for a little while now, and robotic arms have been around much longer. Yet somehow, we don’t have little robot helpers on our desks yet! Thankfully, [Yifei] is working towards that reality with Tabletop Handybot.

What [Yifei] has developed is a robotic arm that accepts voice commands. The robot relies on a Realsense D435 RGB-D camera, which provides color vision with depth information as well. Grounding DINO is used for object detection on the RGB images. Segment Anything and Open3D are used for further processing of the visual and depth data to help the robot understand what it’s looking at. Meanwhile, voice commands are interpreted via OpenAI Whisper, which can feed prompts to ChatGPT for further processing.

[Yifei] demonstrates his robot picking up markers on command, which is a pretty cool demo. With so many modern AI tools available, we’re getting closer to the ideal of robots that can understand and execute on general spoken instructions. This is a great example. We may not be all the way there yet, but perhaps soon. Video after the break.

Continue reading “Tabletop Handybot Is Handy, And Powered By AI” →

Generative AI Hits The Commodore 64

May 15, 2024 by 5 Comments

Image-generating AIs are typically trained on huge arrays of GPUs and require great wads of processing power to run. Meanwhile, [Nick Bild] has managed to get something similar running on a Commodore 64. (via Tom’s Hardware).

A figure generated by [Nick]’s C64. We shall name him… “Sword Guy”!
As you might imagine, [Nick’s] AI image generator isn’t churning out 4K cyberpunk stills dripping in neon. Instead, he aimed at a smaller target, more befitting the Commodore 64 itself. His image generator creates 8×8 game sprites instead.

[Nick’s] model was trained on 100 retro-inspired sprites that he created himself. He did the training phase on a modern computer, so that the Commodore 64 didn’t have to sweat this difficult task on its feeble 6502 CPU. However, it’s more than capable of generating sprites using the model, thanks to some BASIC code that runs off of the training data. Right now, it takes the C64 about 20 minutes to run through 94 iterations to generate a decent sprite.

8×8 sprites are generally simple enough that you don’t need to be an artist to create them. Nonetheless, [Nick] has shown that modern machine learning techniques can be run on slow archaic hardware, even if there is limited utility in doing so. Video after the break.

Continue reading “Generative AI Hits The Commodore 64” →