Position tracking with IMU Sensor and Arduino

Is it possible to track the position of an object in a reference coordinate system with an IMU ( for example MPU 6050)? if yes what is the correct approach in doing that and how can <you define a reference coordinate system?

No. Not possible.

Thanks for your reply
I thought that it could be possible by defining a point as the center of a coordinate system and then finding the position through integration the acceleration.

Is there a way on how to do this? maybe with other sensors

Each sensor has offsets and scaling errors. Those drift over time and temperature. Adding up those errors hundreds of times per second will rapidly accumulate to a massive error.

The gravity vector is impossible to zero. There is no way to distinguish between sitting on the surface of the earth and diving toward the earth at 2G.

Feel free to try but previous experiments have shown it to be a fool's errand.

Your question is about "dead reckoning", which you can look up. It is possible to do this using equipment that costs in the range of $20000-$50000, but not with Arduino.

For indoor positioning or localization, Pozyx works in limited areas. €700.00 to get started.

A Military grade IMU would do the trick.

An Automobile grade IMU with GPS would do the trick.

And the LSM9DS1 is a better option then a MPU605X

I built a INS platform, I use the MPU6050, MPU9050, LSM9DS1, and a MLX90393.

The best performance, from putting the devices through their paces, was from the LSM9DS1 and the MLX90393. For IMU experiments, I recommend the LSM9DS1 over the MPU6050 or MPU9250. I do not recommend a new user to jump onto a MLX90393 until their is some familiarity with using IMU’s.

I ended up writing my own libraries for the MPU9250, LSM9DS1, and the MLX90393 to use the ESP32 SPI bus API.

Fastest data transfers can be had with the LSM9DS1 and the MLX90393. The MLX90393 transfers data the fastest.

I used the LSM9DS1 and the MLX90393 in an auto magnetometer calibration configuration, by wiring and mounting two sensors so their angles of measurement was in opposition.

Unit A mounted on top, Unit B mounted underneath Unit A and position so that X axis, y axis, and z axis are in opposition to each other. I only used the magnetometer measurements from unit B summed with unit A to produce self correcting magnetometer readings.

Yea, you’ll find that after you calibrate a magnetometer all is well until the magnetometer is moved. Now the calibration is out of whack and is no longer calibrated, thus the need for auto correcting.

Idahowalker:
The best performance, from putting the devices through their paces, was from the LSM9DS1 and the MLX90393.

So... How was the performance? Could you put it on a drone and navigate to a specified point and back?

johnwasser:
So... How was the performance? Could you put it on a drone and navigate to a specified point and back?

None of the hobby grade IMU's were sufficient for navigation on an inertial platform.