Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Learning Position From Vehicle Vibration Using an Inertial Measurement Unit

Barak Or, Nimrod Segol, Areej Eweida, Maxim Freydin

TL;DR

This study proposes a vehicle positioning method based on learning the road signature from accelerometer and gyroscope measurements obtained by an inertial measurement unit (IMU) sensor that has a mean error of more than 5 times better for e-scooters and 20 times better for cars.

Abstract

This paper presents a novel approach to vehicle positioning that operates without reliance on the global navigation satellite system (GNSS). Traditional GNSS approaches are vulnerable to interference in certain environments, rendering them unreliable in situations such as urban canyons, under flyovers, or in low reception areas. This study proposes a vehicle positioning method based on learning the road signature from accelerometer and gyroscope measurements obtained by an inertial measurement unit (IMU) sensor. In our approach, the route is divided into segments, each with a distinct signature that the IMU can detect through the vibrations of a vehicle in response to subtle changes in the road surface. The study presents two different data-driven methods for learning the road segment from IMU measurements. One method is based on convolutional neural networks and the other on ensemble random forest applied to handcrafted features. Additionally, the authors present an algorithm to deduce the position of a vehicle in real-time using the learned road segment. The approach was applied in two positioning tasks: (i) a car along a 6[km] route in a dense urban area; (ii) an e-scooter on a 1[km] route that combined road and pavement surfaces. The mean error between the proposed method's position and the ground truth was approximately 50[m] for the car and 30[m] for the e-scooter. Compared to a solution based on time integration of the IMU measurements, the proposed approach has a mean error of more than 5 times better for e-scooters and 20 times better for cars.

Learning Position From Vehicle Vibration Using an Inertial Measurement Unit

TL;DR

This study proposes a vehicle positioning method based on learning the road signature from accelerometer and gyroscope measurements obtained by an inertial measurement unit (IMU) sensor that has a mean error of more than 5 times better for e-scooters and 20 times better for cars.

Abstract

This paper presents a novel approach to vehicle positioning that operates without reliance on the global navigation satellite system (GNSS). Traditional GNSS approaches are vulnerable to interference in certain environments, rendering them unreliable in situations such as urban canyons, under flyovers, or in low reception areas. This study proposes a vehicle positioning method based on learning the road signature from accelerometer and gyroscope measurements obtained by an inertial measurement unit (IMU) sensor. In our approach, the route is divided into segments, each with a distinct signature that the IMU can detect through the vibrations of a vehicle in response to subtle changes in the road surface. The study presents two different data-driven methods for learning the road segment from IMU measurements. One method is based on convolutional neural networks and the other on ensemble random forest applied to handcrafted features. Additionally, the authors present an algorithm to deduce the position of a vehicle in real-time using the learned road segment. The approach was applied in two positioning tasks: (i) a car along a 6[km] route in a dense urban area; (ii) an e-scooter on a 1[km] route that combined road and pavement surfaces. The mean error between the proposed method's position and the ground truth was approximately 50[m] for the car and 30[m] for the e-scooter. Compared to a solution based on time integration of the IMU measurements, the proposed approach has a mean error of more than 5 times better for e-scooters and 20 times better for cars.
Paper Structure (21 sections, 7 equations, 9 figures, 2 tables, 1 algorithm)

This paper contains 21 sections, 7 equations, 9 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related work
  3. Present work
  4. Data Driven based Positioning Inference
  5. Algorithm Description
  6. Tuning the Number of Segments
  7. Data Processing
  8. Road Segmentors
  9. Convolutional neural network
  10. Ensemble random forest
  11. Experiments
  12. Data Collection
  13. Cars
  14. E-scooters
  15. Experimental Results
  16. ...and 6 more sections

Figures (9)

  • Figure 1: An illustration of road segmentation. The proposed methodology is applied for ground vehicle positioning tasks as illustrated. The route is broken into N segments where the IMU measurements are classified into one of the N segments $S_i$, based on unique time-series signal signatures. In this illustration, there is a different road texture in each segment, producing a different and unique IMU signal.
  • Figure 2: Block diagram of the positioning algorithm.
  • Figure 3: The network architecture used for the CNN classifier.
  • Figure 4: (a) The car route (blue) is shown on top of the road map (grey), taken from OpenStreetMap bennett2010openstreetmap., (b) Speed vs. time for four representative recordings with two different cars.
  • Figure 5: (a) The e-scooter route (blue) is shown on top of the road (grey), taken from OpenStreetMap. The route includes on and off-road segments, making for a diverse and challenging route, (b) Speed vs. time for four representative recordings with two different riders and phones.
  • ...and 4 more figures