A boxing band that counts your punches and analyses them as you throw them. Manivannan’s project brings together an Arduino Nesso N1, a Pulse Sensor and a machine learning model trained in Edge Impulse. The built-in 1.14-inch touch display shows the count and changes colour according to heart rate. That way the athlete immediately knows whether they are working in the rest zone or at peak intensity, without looking at an external app.

The Nesso N1 is the board that carries the whole system. The ESP32-C6 microcontroller mounts two 32-bit RISC-V processors, with an HP core up to 160 MHz and an LP core at 20 MHz. Memory amounts to 16 MB of NOR flash and 512 KB of SRAM, figures that leave room for the model and the firmware. On board there is also an SX1262 LoRa module in the 850-960 MHz band, useful if you later want to send data to a second device. The dimensions are 18 mm x 45 mm, so the band stays compact. Power comes from an integrated lithium-polymer battery, with a nominal 5 V input.

From the Nesso_IMU.ino firmware to the four punch classes

The Nesso_IMU.ino firmware initialises the Nesso N1’s IMU, configures the sampling pipeline and sends the readings from the 3-axis accelerometer over the serial port. The serialised motion frames are then acquired by Edge Impulse through the data forwarder. Here the work moves to Edge Impulse Studio, where the frames are segmented and labelled into four classes: Idle, Hook, Jab-Cross and UppeCut.

The preprocessing block is Spectral analysis, the learning block is Classification. The analysis window is 1200 ms and the stride is 750 ms, so the system looks at a little over one second of movement at a time and advances by three quarters of a second. The neural network is trained with 100 training cycles and a learning rate of 0.05. The stated result is 100% accuracy in both the training and the testing phase.

The trained model is exported as an Arduino library and imported into the IDE. From that point on, inference runs locally on the device, with no cloud dependency. The results update the punch count variables, while the Pulse Sensor continuously reads heart rate. Anyone who wants to retrace the path will find the full account of the workflow, from data collection to library export, on Manivannan’s project page on Hackster.io.

The Pulse Sensor on pin GPIO5 and the backlight colours

The pulse sensor must be connected to pin GPIO5, not to the default pin. The PulseSensor Playground library therefore has to be modified: inside PulseSensor.cpp the InputPin variable goes from 7 to 5. The sensor sits on the tip of the index finger and is secured with a hook-and-loop strap, so it stays in contact during the punches.

Heart rate sets the colour of the LED backlight, which becomes an intensity indicator you can read out of the corner of your eye:

• Below 100 bpm the backlight turns green, rest zone or low intensity.
• Between 100 and 140 bpm it turns blue, moderate activity.
• Between 140 and 160 bpm it turns yellow, high intensity.
• Above 160 bpm it turns red, peak intensity.

Rebuilding the band takes only a few parts: the Nesso N1, a Pulse Sensor, hook-and-loop tape and a 5 cm jumper wire kit. Assembly requires no special soldering, because the connections go through the jumpers. What is more, the hook-and-loop strap holds sensor and band together without 3D printing. Anyone looking for a board from the same family for other IoT experiments will find the board with Matter and Bluetooth LE in the catalogue, intended for home automation rather than for sports wearables.

The delicate point is calibrating the heart-rate thresholds. The four ranges on the board hold true for this project, but every athlete has different maximums. So it is worth recording a few rounds and comparing the colours with the real values before trusting the feedback. Sensor position matters too: if it shifts even slightly, the reading wobbles and the colour changes for no good reason.

Punch counting relies on the model’s four classes. Idle increments nothing, while Hook, Jab-Cross and UppeCut push the counter up. Consequently the number on screen is not a simple accelerometer peak, but the result of a classification. The limitation is that the model was trained on one person’s movements: with a different boxer, or with the opposite hand, performance can drop.

Source: https://www.hackster.io/manivannan/smart-ai-boxing-band-bddc39

Related products

• Arduino Nano Matter – IoT Board with Matter® and Bluetooth

The post Boxing Band Counts Punches and Analyses Them in Real Time appeared first on Open Electronics.

A new open-source project demonstrates how to combine computer vision and robotics on a single board. The robot, about the size of a desk, detects a human face through locally executed machine learning and turns toward it in real time, working at roughly 10–15 frames per second. Everything runs on the new Arduino UNO Q, without relying on the cloud.

Dual-processor architecture

The operation leverages the hybrid architecture of the UNO Q, which integrates two distinct units. The Qualcomm MPU microprocessor, running Linux, executes the AI model and the control logic written in Python. The STM32 MCU microcontroller, running Zephyr RTOS, drives the PWM signal of the servomotors with real-time precision.

The two processors communicate via Bridge RPC, a protocol based on MessagePack over an internal serial link, with a round-trip latency of about 8 milliseconds. This way, heavy inference stays separate from the real-time path, keeping motor control deterministic.

Vision, control, and web interface

A USB webcam captures video, while a lightweight face-detection model locates the face in the frame. A proportional controller then converts the horizontal position of the face into differential commands for the wheels. The project uses two ready-made App Lab “bricks”: one for camera-based object detection, the other for the web interface.

In particular, the browser-accessible dashboard lets you monitor detections and adjust steering parameters in real time, without recompiling the code. Every change to the sliders takes effect immediately via Socket.IO, making the tuning phase quick.

Technical details and motion safety

The robot uses differential drive with two continuous-rotation servos: a 1500-microsecond pulse corresponds to stop, while lower or higher values determine rotation in either direction. Additionally, the system limits Bridge calls to a maximum of 20 Hz, because sending commands too quickly would block the serial link.

The project therefore includes several protections: a “coasting” mechanism that maintains the last known position when the face disappears for a moment, a watchdog that stops the robot if detections are interrupted, and an emergency stop that ignores rate limiting to halt the motors immediately. Classified as an intermediate-level project, it is fully documented and released under an open-source license.

The post Face-Following Robot Built on Arduino UNO Q and Edge Impulse appeared first on Open Electronics.

“There goes the neighborhood” isn’t a phrase to be thrown about lightly, but when they build a police station next door to your house, you know things are about to get noisy. Just how bad it’ll be is perhaps a bit subjective, with pleas for relief likely to fall on deaf ears unless you’ve got firm documentation like that provided by this automated noise detection system.

OK, let’s face it — even with objective proof there’s likely nothing that [Christopher Cooper] is going to do about the new crop of sirens going off in his neighborhood. Emergencies require a speedy response, after all, and sirens are perhaps just the price that we pay to live close to each other. That doesn’t mean there’s no reason to monitor the neighborhood noise, though, so [Christopher] got to work. The system uses an Arduino BLE Sense module to detect neighborhood noises and Edge Impulse to classify the sounds. An ESP32 does most of the heavy lifting, including running the UI on a nice little TFT touchscreen.

When a siren-like sound is detected, the sensor records the event and tries to classify the type of siren — fire, police, or ambulance. You can also manually classify sounds the system fails to understand, and export a summary of events to an SD card. If your neighborhood noise problems tend more to barking dogs or early-morning leaf blowers, no problem — you can easily train different models.

While we can’t say that this will help keep the peace in his neighborhood, we really like the way this one came out. We’ve seen the BLE Sense and Edge Impulse team up before, too, for everything from tuning a bike suspension to calming a nervous dog.