Golf Putter Smoothness Analysis

Other Hardware Sensors
1

Good day community :smiley: . I am still new to developing stuffs and electronics projects, but to be specific currently trying projects with mini-boards, but seeking for people with a good understanding of building motion sensors, for collecting swing data from the golf putter.

I am on to building a very simple LED 5*5 matrix, coupled by ESP32 and IMU6050 sensor, and the setup casing mounted on a golf putter shaft to collect data such as club: Face Angles, Time, Speed & Rotation, basically to analyze the smoothness of the putter stroke, by giving out results, through the LED lights, where the center array will be green indicating smooth/ideal putter stroke, while the left and right, 2 rows for each side, left side from the center being jerky/fast/jabbed, and the other side being too short/stiff/flinch.

image
image1041×697 46.3 KB

Above, is the simple connection of the circuit, and at later I did add a LiPo 3.8v battery to power the ESP board without USB power, and a mini- switch. And pls, the LED VCC is from 3.3v of the ESP32 not the 5v as seen in the picture.

Through various researches done, I still need to understand, how to make this work in its simplest design and setup,

  1. what should be really considered to have a full functional setup?,
  2. do I have to do calibration of the sensor, every time I want to start hit the golf ball
  3. are there any golf-specific logic for stroke analysis, that I should include or something?
  4. are there parameters that i should put into consideration?
  5. how to make the sensor to be very sensitive to slow motion during the swing from back to froward stroke, and also reduce lagging, or receiving unwanted blinks of the LED, that I still do not understand the cause.
  6. what could be the proper orientation of this kind of sensor, for getting acceleration and rotation readings proper in relation to the position of the golf putter shaft.
  7. not sure if I'm missing anything, but kindly if there are things that can help my understanding, I will be glad to be guided, including the right forums I should visit more for this kind of questions,

Thank you very much.

2

I don't play golf, so I can't comment on the golf-specific aspects. However, from a purely analytical and engineering perspective, I see a few points that may be worth considering.

Your hardware choice is perfectly reasonable for a first prototype. I wouldn't worry too much about sensor orientation, calibration or the LED output just yet.

The difficult part is not reading the MPU6050 - it's deciding what actually defines a "good" putting stroke.

Before writing the analysis code, I would:

  • Record raw accelerometer and gyroscope data from many strokes.
  • Save the data to a PC (CSV is sufficient).
  • Plot the curves.
  • Compare good and bad putts manually.

Only after you can clearly identify which features distinguish the strokes should you implement an algorithm and map the result to your LED matrix.

Otherwise you'll probably end up changing thresholds blindly.

I'd also recommend separating the project into three independent parts:

  1. Reliable sensor acquisition.
  2. Stroke analysis.
  3. LED visualization.

Getting each part working independently makes debugging much easier.

One final thought: I wouldn't use the LED matrix while developing the analysis itself. Log the raw data to a PC first. The LEDs are an excellent way to present the final result, but not the best tool for developing the classifier.

3

Thank you so much, for your thoughts and important explanations of things I should be considering, I am a very junior mechanical engineer, and with you, digging into the engineering perspectives, it is surely helpful, for my progress and way forward, i should put into practice these and also deepen into the direction of the thought you have suggested me.
My kind regards.