Offroad Robotics

Permanent URI for this collection

Offroad Robotics is a multidisciplinary research group at Queen’s University. Our researchers have expertise in mining, mechanical, and electrical & computer engineering, and are passionate about field robotics, mechatronics, and systems control.

This community includes research outputs produced by faculty and students. Submitting works to QSpace may enable compliance with the Tri-Agency Open Access Policy on Publications.

When you submit your work to QSpace, you retain copyright and grant the Library a non-exclusive license to distribute and preserve. Works are open access unless restricted by the creator.

Browse

Recent Submissions

Now showing 1 - 5 of 39
  • Item
    Monkey see, monkey do: Constant time delay leader following for wheeled mobile robots using uncertainty-tuned model predictive control
    (IEEE, 2024-04-15) Wang, Kai; Givigi, Sidney; Marshall, Joshua A.
    This paper presents a constant time delay leader-follower system for wheeled mobile robots that can be used in a military vehicle convoy. It allows a follower vehicle to follow the same trajectory travelled by a leader with a specified time delay by estimating a sequence of waypoints travelled by the leader. This paper focuses on the design of the two major components of the leader-follower system: the estimator and the controller. Different from previous works on constant time delay leader-follower systems, where the two components are designed relatively independently, we propose a design that connects the two components by having a model predictive controller (MPC) whose tuning is based on the uncertainties of the estimated leader waypoints. By weighting the waypoint states in the MPC cost function as inversely proportional to their uncertainties, we achieve better tracking performance under the condition of noisy odometry sensor data compared to the traditional practice equally-weighted state errors. This approach reduces the dependence on empirical tuning and performance was studied both in simulation and on real robots in representative off-road environments. Currently, the system is designed to track the linear, longitudinal trajectories of the leader, as a starting point for tracking nonlinear trajectories in future work.
  • Item
    Touchless control of heavy equipment using low-cost hand gesture recognition
    (IEEE, 2022-03-01) Khaleghi, Leyla; Artan, Unal; Etemad, Ali; Marshall, Joshua
    Human-machine interaction using remote hand gestures is becoming increasingly prevalent across various industries. However, their potential application to heavy construction equipment is often overlooked. This paper presents a robust and inexpensive hand gesture recognition system that was implemented and tested on a robotic 1-tonne wheel loader. The system uses an RGB camera paired with a laptop to process, in real time, hand gestures to control the loader. We first design 4 unique gestures for controlling the loader and then collect 26000 images to train and test a neural network for hand gesture recognition. Our system uses robust landmark detection using an off-the-shelf system prior to gesture recognition. We successfully controlled the loader to excavate in a rock pile by using the proposed hand gesture recognition system.
  • Item
    Mapping waves with an uncrewed surface vessel via Gaussian process regression
    (IEEE, 2023-05-29) Sears, Thomas M. C.; Cooper, M. Riley; Marshall, Joshua A.
    Mobile robots are well suited for environmental surveys because they can travel to any area of interest and react to observations without the need for pre-existing infras- tructure or significant setup time. However, vehicle motion constraints limit where and when measurements occur. This is challenging for a single vehicle observing a time-varying phenomenon, such as coastal waves, but the ability to generate a spatiotemporal map would have immediate scientific and engineering applications. In this paper, an uncrewed surface vessel (USV) was used to measure waves on the coast of Lake Ontario, Canada. Data were collected from a low-cost inertial measurement system onboard the USV and processed in an offline Gaussian process regression (GPR) workflow to create a spatiotemporal wave model. Frequency analysis of raw sensor data was used to best select and design kernel functions, and to initialize hyperparameters. The relative speed of the waves limited the ability to make complete wave reconstructions, but GPR captured the dominant periodic components of the waves despite irregularities in the signals. After optimization, the hyperparameters indicate a dominant signal with a wave period of 0.87 s, which concurs with ground truth estimates.
  • Item
    Towards unsupervised filtering of millimetre-wave radar returns for autonomous vehicle road following
    (IEEE, 2023-03-01) Sacoransky, Dean; Marshall, Joshua A.; Hashtrudi-Zaad, Keyvan
    Path planning and localization in low-light and inclement weather conditions are critical problems facing autonomous vehicle systems. Our proposed method applies a single modality, millimetre-wave radar perception system for the detection of roadside retro-reflectors. Radar-based perception tasks can be challenging to perform due to the sparse and noisy nature of radar data. We propose the use of an unsupervised learning approach for filtering radar point clouds through Density-Based Spatial Clustering of Applica- tions with Noise (DBSCAN). The DBSCAN algorithm segments retro-reflector points from noise points, thus providing the autonomous vehicle with a predicted path for the road ahead. We tested the approach via indoor experiments that make use of Continental’s ARS 408 radar, a mobile Husky A2000 robot, and a Vicon motion capture system for ground truth validation. The experimental results of the proposed system demonstrated a classification accuracy of 84.13 % and F1 score of 83.71 %.
  • Item
    Mapping of spatiotemporal scalar fields by mobile robots using Gaussian process regression
    (IEEE, 2022-10) Sears, Thomas M C; Marshall, Joshua A
    Spatiotemporal maps are data-driven estimates of time changing phenomena. For environmental science, rather than collect data from an array of static sensors, a mobile sensor platform could reduce setup time and cost, maintain flexibility to be deployed to any area of interest, and provide active feedback during observations. While promising, mapping is challenging with mobile sensors because vehicle constraints limit not only where, but also when observations can be made. By assuming spatial and temporal correlations in the data through kernel functions, this paper uses Gaussian process regression (GPR) to generate a maximum likelihood estimate of the phenomenon while also tracking the estimate uncertainty. Spatiotemporal mapping by GPR is simulated for a single fixed- path mobile robot observing a latent spatiotemporal scalar field. The learned spatiotemporal map captures the structure of the latent scalar field with the largest uncertainties in areas the robot never visited.