Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RTS-GT: Robotic Total Stations Ground Truthing dataset

Maxime Vaidis, Mohsen Hassanzadeh Shahraji, Effie Daum, William Dubois, Philippe Giguère, François Pomerleau

TL;DR

The proposed Robotic Total Stations Ground Truthing dataset (RTS-GT) dataset is proposed to support localization research with the generation of six-Degrees Of Freedom (DOF) ground truth trajectories, making it the most extensive dataset of RTS-based measurements to date.

Abstract

Numerous datasets and benchmarks exist to assess and compare Simultaneous Localization and Mapping (SLAM) algorithms. Nevertheless, their precision must follow the rate at which SLAM algorithms improved in recent years. Moreover, current datasets fall short of comprehensive data-collection protocol for reproducibility and the evaluation of the precision or accuracy of the recorded trajectories. With this objective in mind, we proposed the Robotic Total Stations Ground Truthing dataset (RTS-GT) dataset to support localization research with the generation of six-Degrees Of Freedom (DOF) ground truth trajectories. This novel dataset includes six-DOF ground truth trajectories generated using a system of three Robotic Total Stations (RTSs) tracking moving robotic platforms. Furthermore, we compare the performance of the RTS-based system to a Global Navigation Satellite System (GNSS)-based setup. The dataset comprises around sixty experiments conducted in various conditions over a period of 17 months, and encompasses over 49 kilometers of trajectories, making it the most extensive dataset of RTS-based measurements to date. Additionally, we provide the precision of all poses for each experiment, a feature not found in the current state-of-the-art datasets. Our results demonstrate that RTSs provide measurements that are 22 times more stable than GNSS in various environmental settings, making them a valuable resource for SLAM benchmark development.

RTS-GT: Robotic Total Stations Ground Truthing dataset

TL;DR

The proposed Robotic Total Stations Ground Truthing dataset (RTS-GT) dataset is proposed to support localization research with the generation of six-Degrees Of Freedom (DOF) ground truth trajectories, making it the most extensive dataset of RTS-based measurements to date.

Abstract

Numerous datasets and benchmarks exist to assess and compare Simultaneous Localization and Mapping (SLAM) algorithms. Nevertheless, their precision must follow the rate at which SLAM algorithms improved in recent years. Moreover, current datasets fall short of comprehensive data-collection protocol for reproducibility and the evaluation of the precision or accuracy of the recorded trajectories. With this objective in mind, we proposed the Robotic Total Stations Ground Truthing dataset (RTS-GT) dataset to support localization research with the generation of six-Degrees Of Freedom (DOF) ground truth trajectories. This novel dataset includes six-DOF ground truth trajectories generated using a system of three Robotic Total Stations (RTSs) tracking moving robotic platforms. Furthermore, we compare the performance of the RTS-based system to a Global Navigation Satellite System (GNSS)-based setup. The dataset comprises around sixty experiments conducted in various conditions over a period of 17 months, and encompasses over 49 kilometers of trajectories, making it the most extensive dataset of RTS-based measurements to date. Additionally, we provide the precision of all poses for each experiment, a feature not found in the current state-of-the-art datasets. Our results demonstrate that RTSs provide measurements that are 22 times more stable than GNSS in various environmental settings, making them a valuable resource for SLAM benchmark development.
Paper Structure (12 sections, 5 figures, 3 tables)

This paper contains 12 sections, 5 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. Hardware
  4. Data collection
  5. Environments
  6. Ground truth protocol
  7. Calibration and time synchronization
  8. Data format
  9. Discussion and challenges
  10. Precision and reproducibility
  11. Challenges encountered
  12. Conclusion

Figures (5)

  • Figure 1: Example of two areas provided in the dataset (bird's-eye view). Left: a forest. Right: a park on a campus. Maps in black are based on lidar scans, while the colored spheres represent the scaled-up uncertainty of the provided ground truth in millimeters. A Clearpath Warthog platform was used during these experiments, whose paths are illustrated in green.
  • Figure 2: Setup used in the tunnel sites. Left: RTS setup with the HD2 robot. Right: one deployment done in a 120 tunnel. Because the floor was slippery, heavy weights were added to stabilize the RTS tripods.
  • Figure 3: Setup used on the campus with the Warthog UGV. A GNSS fixed base station sends corrections to the three GNSS rovers on the robot. Three prisms are tracked by three RTS. Data are collected by three Raspberry Pi clients connected by USB to the RTS. A LoRa communication protocol is used to send data to a Raspberry Pi master located on the UGV. The lidar and IMU are on the front part of the Warthog.
  • Figure 4: Distribution of errors arising for the two setups. Left: inter-prism and inter-GNSS distances. Right: inter-precision distances. Results obtained from the RTS are denoted in blue, whereas those from the GNSS are shown in orange. The median error values are in the center of each box, and the IQR is indicated alongside for reference.
  • Figure 5: Issues encountered while collecting the dataset. Left: difficulties of leveling on deep snow. Middle: dust on a lens which was interfering with the tracking mode of the RTS. Right: supports added on the Warthog to reduce vibrations on the top prism.