Inside the Triki motion controller and a bit of reverse engineering

gulson wrote:

Thanks for sharing the mods! There’s no shame in using AI – if you’ve got a hammer, you’ve got to use it! A really cool YouTube channel!
Who knows, maybe thanks to the amphibian we’ll instil a bit of ele in the nation...

A direct answer to the question

These tricks can be viewed not as a ‘game controller’, but as a ready-made, cheap wireless BLE module with an IMU motion sensor, an LED and a button or buttons . In practice, this means: a small gesture controller, a beacon, a vibration sensor, an IoT remote control or an educational module.

The most interesting practical applications:

- a gesture remote control for computers, presentations and multimedia ,
- smart home controller via BLE/MQTT/Home Assistant ,
- shock, tilt and drawer/door-opening sensor ,
- assistive interface for people with limited mobility ,
- simple activity, rehabilitation and posture sensor ,
- wireless vibration sensor for machinery and tools ,
- controller for robots, RC models, vehicles and interactive installations ,
- educational tool for learning about BLE, MEMS, signal filtering and IoT ,
- a business gadget for events, exhibitions, escape rooms, training sessions and gamification .

The greatest value lies in the fact that the hardware is already complete: the casing, battery, BLE, antenna, nRF52810 microcontroller, motion sensor and LED. All you need to do is write the software layer to receive data via the Nordic UART Service.

---

Detailed analysis of the problem

What Triki actually has

Based on the information available so far in this thread, it appears that Triki features:

- Nordic nRF52810
An ARM Cortex-M4 microcontroller with Bluetooth Low Energy. Suitable for small IoT devices, remote controls, beacons, sensors and simple HID/BLE interfaces.

- SF248 MEMS sensor
Most likely an IMU-type chip, i.e. an accelerometer + gyroscope. Since we’ve managed to obtain data from both the gyroscope and the accelerometer, we effectively have a 6-axis motion sensor.

- BLE communication via Nordic UART Service (NUS)
This is very convenient, as on the PC, Android, Raspberry Pi or ESP32 side, the transmission can be treated almost like a simple serial port over BLE.

- Software-controlled LED
It may seem like a minor detail, but in practical applications it provides very useful feedback: connection status, an alarm, gesture confirmation, low battery, etc.

- SWD pads on the PCB
If the pinout is confirmed and there is no read/write lock, it may be possible to flash your own firmware. But even without that, simply using the factory firmware via NUS already offers plenty of possibilities.

---

1. Gesture remote control for a computer

This is one of the simplest and most useful applications.

Examples

- Presentation remote control
- turn clockwise: next slide,
- turn anti-clockwise: previous slide,
- shake: black screen / pause,
- click: confirm.

- Air mouse
- tilt along the X-axis: move the cursor left/right,
- tilt along the Y-axis: move the cursor up/down,
- button: click,
- quick gesture: scroll.

- Multimedia controller
- tilt right/left: next/previous track,
- tilt up/down: volume,
- tap: play/pause.

How to do this technically

Diagram:



On a PC, you can use, for example:

- Python +

bleak

for BLE,
-

pyautogui

/

pynput

for keyboard and mouse emulation,
- a browser-based application using the Web Bluetooth API,
- a small programme in C# on Windows.

This is a good first demonstration project, as it requires no additional electronics.

---

2. Smart home control

This can be a very handy remote control for Home Assistant or your own automation system.

Example gestures



Architecture



The nRF52810 does not have Wi-Fi, so a gateway is required to connect to the IP world:

- Raspberry Pi,
- laptop/mini-PC,
- ESP32 with BLE and Wi-Fi,
- Android phone as an intermediary,
- web browser with Web Bluetooth.

Practical example

Triki is lying by the bed. A single shake turns the lights off, a twist adjusts the brightness of the bedside lamp, and holding it down or tilting it for longer activates the ‘night’ scene.

This is much more natural than searching for your phone and launching an app.

---

3. Opening, movement and impact sensors

The accelerometer allows you to turn Triki into a simple wireless intrusion sensor.

Possible applications

- a drawer containing documents,
- a toolbox,
- a cellar door,
- a window,
- a bicycle in the garage,
- a suitcase,
- a parcel during transport.

Operating logic



Example criteria:

- sudden surge in acceleration,
- change in angle relative to gravity,
- series of vibrations,
- movement following a long period of inactivity.

This is no substitute for a certified alarm, but as a home or workshop project it is very interesting.

---

4. Support for the elderly and those with limited mobility

This is where Triki can have some really useful applications.

Possible functions

- Help request button/gesture
A shake or a specific gesture sends a notification to the carer.

- Fall detection
A sudden acceleration, followed by a lack of movement for, say, 30–60 seconds, may trigger an alarm.

- Alternative computer control interface
For someone who cannot use a mouse comfortably, Triki, when attached to the hand, head or armrest, can control the cursor or trigger simple commands.

- Remote control for the bed, lamp, TV and blinds
One controller and a few simple gestures.

Important disclaimer

Common sense must be exercised when using this for care or medical purposes. As a domestic aid, a prototype or an additional alert system — by all means. As the sole safety system for a sick person — no, as the device is not certified for this purpose.

---

5. Rehabilitation, activity, exercises

The IMU enables the detection of repetitive movements. This is very well suited to simple rehabilitation exercises.

Examples

- wrist exercise repetition counter,
- range of motion monitoring,
- simple rehabilitation game,
- balance exercises,
- movement pace control,
- detection of excessively sudden movements.

Example:



This could be a great educational project with real-world applications: electronics + programming + biomechanics + gamification.

---

6. Posture monitor

The tags can be attached to a harness, T-shirt, wristband or rucksack and used as a tilt sensor.

Example of how it works

1. The user sits up straight.
2. The app records the reference position.
3. If the tilt angle exceeds, say, 15–20° for a prolonged period, an alert is triggered.
4. The LED flashes or the phone emits a quiet beep.

You don’t even need a gyroscope here — the accelerometer alone, which measures the direction of the gravitational vector, is sufficient to detect static tilt.

---

7. Wireless vibration sensor for workshops and small businesses

This is a very interesting ‘semi-industrial’ application.

Applications

- monitoring the operation of a motor,
- a pump,
- a fan,
- a compressor,
- a grinder,
- a bench drill,
- a 3D printer,
- a generator,
- a washing machine or tumble dryer in a domestic setting.

What can be measured

- whether the device is operating,
- how long it has been operating,
- whether vibrations have increased compared to normal levels,
- whether an impact has occurred,
- whether the machine has been moved,
- whether an unusual vibration pattern is present.

Business example:



Of course, professional vibration diagnostics require the appropriate bandwidth, sampling frequency, calibration and sensor mounting. But for simple ‘working/not working’ or ‘vibrating more than usual’ detection, such a gadget may suffice.

---

8. Overload and transport logger

Following firmware modifications or when transmitting continuously to the Triki receiver, it can be used as an impact logger.

Examples

- whether a parcel was thrown,
- whether a crate was knocked over,
- how hard a bike or scooter braked,
- whether equipment in a car shifted during transport,
- recording impacts on a suitcase.

The limitations here will be memory, battery life and access to data. If the factory firmware only streams data, a receiver will be needed nearby. If it were possible to flash custom firmware, a local event buffer could be created and the data read later via BLE.

---

9. Controller for robotics and RC models

The Triki is ideally suited as a simple wireless controller.

Examples

- RC car:
- tilt forwards/backwards: accelerator/brake,
- tilt left/right: steering,

- mobile robot:
- forward gesture: drive,
- wrist rotation: turn,
- shake: emergency stop,

- robot arm:
- gyroscope controls the rotation axis,
- accelerometer controls the position,
- button activates the gripper.

Typical setup



When it comes to models and robotics, you need to bear in mind delays and reliability. For fun and BLE demonstrators, this is sufficient. For critical applications, it is better to have an additional physical ‘stop’ switch.

---

10. Interactive exhibitions, events and marketing

There are very real business applications here.

Examples

- a trade fair stand where a customer rotates a physical prop and a 3D model rotates on the screen,
- gesture voting during a presentation,
- an interactive game at an event,
- an escape room: you have to perform the correct sequence of movements,
- a museum: shake or rotate an item to trigger an animation,
- a shop: the controller as a cheap ‘magic token’ to activate content.

From an event company’s perspective, this is great because the equipment is cheap, compact and requires no cables. And for the client, it looks impressive.

---

11. BLE beacon with additional motion data

The nRF52 is very well suited to the role of BLE beacons. It can act as an identifier, but one enhanced with motion data.

Examples

- locating tools in a workshop,
- identifying a person at a training course,
- an attendance token at an event,
- a stock tag in a warehouse,
- confirming task completion with a gesture,
- detecting when an item is taken out of a zone.

Diagram:



This will not provide precise positioning like UWB, but it is sufficient for determining whether an item ‘is in this zone’ or ‘has left the zone’.

---

12. MIDI controller and musical applications

A very interesting, less obvious direction.

Examples

- a hand movement changes the synthesiser’s filter,
- tilting controls the delay/reverb effect,
- shaking triggers a sample,
- rotating changes the pan,
- a button starts a loop.

Schematic:



This could be a great project for electronic musicians. A ‘gesture controller’ like this normally costs considerably more than a promotional gadget.

---

13. Education: a very good kit for learning electronics

Here, the ‘Amphibian’ could actually ‘instil an interest in electronics in the nation’.

With just one small device, you can teach:

- Bluetooth Low Energy,
- Nordic UART Service,
- protocol reverse engineering,
- Bluetooth login from Android,
- frame analysis in Wireshark,
- working with an IMU,
- signal filtering,
- gesture detection,
- the basics of kinematics,
- communication with a browser via Web Bluetooth,
- integration with Python, Node-RED, MQTT and Home Assistant,
- microcontroller programming via SWD, if you manage to delve deeper.

This is a very rewarding topic for a series of workshops:

1. ‘Read data from Triki’.
2. ‘Plot an accelerometer graph’.
3. “Detect a shake”.
4. “Make a remote control for slides”.
5. “Connect to Home Assistant”.
6. “Control a robot”.
7. “Write your own firmware”.

---

Proposed library / SDK for the community

The greatest progress would come from creating a simple library that hides the details of BLE and frames.

Useful functions



A sensible data structure



It would be a good idea to prepare versions for at least:

- JavaScript / Web Bluetooth,
- Python / Bleak,
- ESP32 / Arduino or ESP-IDF,
- possibly C# for Windows.

---

Gesture detection — a practical approach

You don’t need any artificial intelligence to get started. Simple thresholds and filtering are enough.

Tilt

The angle can be estimated using an accelerometer:



This is sufficient for controlling a lamp, a blind or a presentation.

Shake

We calculate the acceleration module:



If

a

deviates sharply from 1 g and there are several changes in direction within a short time, we have a shake.

Tap

A short acceleration pulse, e.g.:



Idle

Low variance in the accelerometer and gyroscope for a few seconds:



Rotation

Here we use the gyroscope:



Only later can you experiment with a complementary filter, a Kalman filter or gesture classification using machine learning.

---

Limitations worth bearing in mind

1. The exact layout of the SF248 is unknown

As there is no data sheet, the following must be determined experimentally:

- the accelerometer’s range, e.g. ±2 g, ±4 g, ±8 g,
- the gyroscope’s range, e.g. ±250/500/1000/2000 °/s,
- the sampling frequency,
- data scaling,
- byte order,
- offsets,
- noise,
- gyroscope drift.

Without this information, the measurements will be rather approximate.

2. Gyroscope drift

After integration, the gyroscope ‘drifts’. It’s good for short-term orientation, but the error accumulates over time. That’s why, to determine tilt reliably, it needs to be combined with an accelerometer.

3. BLE is not a cable

BLE is energy-efficient, but:

- it has delays,
- it can lose packets,
- it depends on the phone/computer,
- its range depends on the casing, the antenna and the surroundings.

For controlling a lamp — OK. For an emergency stop on a large machine — not as the sole communication channel.

4. Battery

If Triki is constantly streaming IMU data at a high frequency, the battery will drain much faster than in beacon/standby mode. For home sensors, it is worth using the following mode:



However, this requires firmware support or clever use of what is already in the factory software.

5. Legal issues and security

For hobbyist use — complete educational freedom. For commercial implementation, you must be mindful of:

- trademarks and branding,
- terms of use for the application/protocol,
- device certification following firmware modification,
- BLE data security,
- medical and industrial applications, where standards and responsibility for system operation are required.

---

The best projects ‘to get you started’

If I had to choose a few projects that best demonstrate Triki’s potential, I would suggest:

Project	Difficulty	Practical effect
Accelerometer and gyroscope graph	low	education, diagnostics
Presentation remote control	low	practically useful
Gesture-based volume control	low	instant ‘wow’ effect
Drawer/door alarm	medium	home use
Integration with Home Assistant	average	home automation
Exercise repetition counter	average	rehabilitation/sport
Controlling a toy car with ESP32	average	robotics
Machine vibration sensor	medium/high	workshop use
MIDI controller	medium	music/installations
Custom firmware via SWD	high	full hardware control

---

My suggestion for the direction of this thread

The most valuable thing for the community would be to create a package:

1. Description of the NUS Triki protocol
Commands, responses, frame formats, sample logs.

2. Python and JS library
A simple API for reading data.

3. A few ready-made examples
- presentation remote control,
- 3D orientation chart,
- LED control,
- motion alarm,
- MQTT for Home Assistant.

4. Table of known hardware versions
- PCB designation,
- IMU chip designation,
- firmware version,
- any differences in the bezels.

5. Description of SWD pins
- VCC,
- GND,
- SWDIO,
- SWDCLK,
- RESET, if available.

This turns the toy into a community-driven mini-ecosystem: ‘SDK Tricks’. And this could make for a really cool educational project on Elektroda.

---

Summary

Triki is a good example of a device which, from a marketing perspective, is a ‘game controller’, but technically is a fully-fledged, small IoT module with BLE and an IMU. I see the greatest potential in three areas:

1. Education and popularisation of electronics — BLE, sensors, reverse engineering, programming.
2. Practical domestic and assistive applications — gesture-controlled remote, alarm, support for the elderly, rehabilitation.
3. Business prototypes — events, interactive exhibitions, vibration monitoring, logistics, gamification.

If the thread’s author were to make available a library for interpreting data from Triki, the barrier to entry would be significantly lowered. Then even those with no experience of BLE would be able to create their own application in a few dozen lines of code.