Student Projects

A list of available projects for seniors or students seeking independent research opportunities can be found on the student projects page. These projects cover a broad range of research topics in robotics. These are meant to be a general topic for a research project, with the specifics determined by an individual's knowledge, expertise and interests.

Long-Term Autonomy and Persistent Navigation in Spatiotemporally Dynamic Environments

The goals of this project are to overcome the theoretical and technical challenges of developing a general prediction, control and planning framework for autonomous navigation and sampling of dynamic ocean features. The PI will make contributions to the areas of navigation, deliberation, prediction, and targeted sampling to extend autonomy in marine robotics. In pursuit of this goal, the objectives of this proposal are to develop the broad principles that enable autonomous aquatic vehicles to 1) improve navigation and localization for operation in dynamically- evolving environments, 2) demonstrate the utility of prior information in planning and deliberation, 3) investigate human in-the-loop control strategies and the associated data analysis and transmission necessary, and 4) experimentally validate research outcomes through simulation, laboratory tests, and field trials.
This research is being carried out through an international fellowship provided by the Queensland Smart State Program at the Monterey Bay Aquarium Research Institute in Moss Landing, CA USA.

Funding Source(s): 
Office of Naval Research Grant Number N00141612634, Program Manager: Jason Stack


Putting Sensors in the Right Place at the Right Time

This research is investigating and developing observational strategies to be employed by Autonomous Underwater Vehicles to effectively sample dynamic, coastal ocean processes, and monitor freshwater reserves. The need to understand dynamic coastal ocean biogeochemical processes, which have poor computational analogs, has highlighted the need for novel and mixed-initiative robotic control. Automated and efficient sampling and monitoring of freshwater reserves provides a necessary assessment of the quality and quantity of Queensland's drinking water supply. With single or multiple vehicles in any body of water, one key, science-driven issue is to figure out when and where to sample. The sampling strategies developed will be implemented onto an AUV owned by QUT, specifically designed for environmental monitoring and aquatic sampling. Gathered data will be provided to end users (SEQ Water, CSIRO, Great Barrier Reef Marine Park Authority) to answer specific science-driven questions about dynamic processes; variability in ocean pH, greenhouse gas content, quality of fresh water reserves.
This research is being carried out through an international fellowship provided by the Queensland Smart State Program at the Monterey Bay Aquarium Research Institute in Moss Landing, CA USA.

Funding Source(s): 
Queensland Government (Smart State Program), QUT ECARD Program


Crowd-Sourced Control of Autonomous Marine Vehicles

Persistent platforms like the Liquid Robotics Wave Glider and Webb Slocum Underwater Glider are changing the way ocean science is conducted. These platforms are ideal for education and outreach, teaching children about the multiple elements of large
science experiments (similar to controlling the Mars Rover). Additionally, these persistent platforms enable interesting research in the democratization of large-scale experiments, e.g., the platforms have physical constraints with an objective to accomplish multiple (and potentially conflicting) science goals. How do the vehicles focus on the issues important to many?
The basic idea of this project is relatively simple – utilise crowd-sourced input to control vehicles in the ocean. Users vote on a path or direction of travel. In general, crowd-sourcing is the act of outsourcing tasks, traditionally performed by an employee or contractor, to an undefined, large group of people or community, a crowd through an open call. Here, the crowd-sourced input provides the available options, and the on-board autonomy system decides what is best to execute, given the current operational constraints, e.g., hazard mitigation, collision avoidance, vehicle speed, etc. The voters do not get total control, but only supply the vehicle with options for its next move. The crowd uses a web-based user interface to visualize the current location and path of the vehicle, e.g., center of Monterey Bay, heading North at 1 m/s. Additionally, relevant science data from satellites, buoys, etc. can be overlaid on the map to inform the users of the latest state of the ocean. Based on this information, the crowd chooses the next direction for the vehicle to be executed.
Each user can access the website on their internet-connected device, where they can see the current location of the platform, and can vote for one of six choices (Unsure, Stop, North, South, East, or West). The server tallies all votes every 10 seconds, determines the winning command, validates this command, and updates the vehicle state (i.e., sends the command to vehicle).
To try your hand at controlling a vehicle in the ocean, please visit the voting website. This allows users to control a Wave Glider platform in Monterey Bay, California USA.

Funding Source(s):



Control of Minimally-Actuated Aquatic Vehicles

The primary premise of this research is to consider the interesting question: Can you control the trajectory of a profiling float?

Establishing a persistent presence in the ocean with an Autonomous Underwater Vehicle capable of observing temporal variability of large-scale ocean processes requires a unique sensor platform. In this research, we examine the utility of Lagrangian profiling floats for such extended deployments. We are investigating strategies that utilise ocean model predictions to facilitate a basic level of autonomy to achieve general control of this minimally-actuated underwater vehicle. We seek to extend experimentally validated techniques for using ocean current models to control under-actuated autonomous underwater vehicles to the application of these methods on profiling floats. With the appropriate vertical actuation, and utilising spatiotemporal variations in water speed and direction, we investigate the level of controllability that can be achieved.

Funding Source(s): Office of Naval Research - Global


Autonomous Underwater Assessment of Change in Coral Reef Health

This project develops innovative robotics solutions for assessing large- scale responses of coral reef ecosystems to climate change. This is accomplished through a unique blend of persistent underwater navigation, vision-based sensing and cooperative mission planning. Coral reefs are a critical part of the marine ecosystem and support a rich diversity of life with consequent economic value and social amenity. However, sea surface temperatures have increased over the past few decades, resulting in widespread coral bleaching at an ever-increasing rate. Projected increases in global temperatures of 2 - 4.5°C by 2100 indicate that mass coral bleaching events are likely to become an annual phenomena by 2050. The widespread mortality of corals following mass bleaching events reduces the structural complexity of reefs – eliminating the three dimensional habitat, which is critical to maintaining diversity and population of coral reef fish communities. Despite the importance of coral reef ecosystems, spatial and temporal dynamics of coral bleaching events are poorly understood. Most surveys of coral bleaching have been conducted using either human divers or remote sensing (satellite imagery), but these approaches are fundamentally limited in scale. Satellites are able to cover large spatial scales (10s of m2 to km2), and approximate percentage of coral cover, but are unable to resolve fine-scale ecological changes (centimeters to meters) of individual corals. In situ measurement by divers can provide this data, but at limited spatial scales (<1 km2). Robotics provides a novel solution to the fundamental problem of large- scale area coverage, and represents a viable approach to quantifying the extent of fine-scale coral bleaching and reef structural complexity over large areas (>100 km2), for long periods of time (days to weeks), and importantly, at comparatively low cost. The application of our unique combination of vision and planning strategies on autonomous underwater vehicles will enable a synoptic view of coral reef ecosystems at an unprecedented spatial resolution. There are three critical research problems in better understanding coral reef health: 1) monitoring of individual coral with respect to bleaching and structural complexity across large spatial scales (>100 km2), 2) quantifying changes in coral bleaching and structural complexity over time, 3) using these measures to assess reef and reef ecosystem health and to predict health under various scenarios. This robotics project will deliver innovative solutions to the first two problems, while providing the data necessary to address the third problem. We address problems 1 and 2 through the following Aims.
AIM 1: Develop motion planning and control techniques that enable large scale spatial coverage by a fleet of persistent, free floating, autonomous robots controlled with novel distributed planning strategies based on ocean model predictions.
AIM 2: Develop sensor-based, mission-adaptation techniques that enable robots to respond to the environment and maximize visual and 3D structural information via an active choice of new viewpoints, subject to vehicle motion constraints.
The data collected across multiple spatial resolutions in Aims 1 & 2 will be unprecedented in scale, and resolution and have a transformative impact on the conduct of coral reef ecology - quantifying the response of coral reef ecosystems to climate change.

Funding Source(s):


Persistent and Adaptive Ocean Observation and Monitoring

Ocean processes are dynamic, complex, and occur on multiple spatial and temporal scales. This project aims to obtain a synoptic view of these processes by developing algorithms that produce persistent monitoring missions for underwater vehicles. Balancing path following accuracy and sampling resolution in a given region, the developed techniques provide data collection capabilities across a range of spatiotemporal resolutions. We address a pressing need among ocean scientists to efficiently and effectively gather high-value data in key areas of interest. Developed algorithms are implemented in sea trials on underwater vehicles in southern California, and in Monterey Bay, California.

Funding Source(s): Office of Naval Research, National Science Foundation, QUT ECARD Program


Path Planning for Autonomous Underwater Vehicles

Intelligent path planning for AUVs is required to manoeuvre a vehicle to high-valued locations to perform data collection. In this research area, we focus on the use of ocean models to:
    1. Track evolving features of interest
    2. Increase navigational accuracy of AUVs
    3. Optimise sampling strategies to place the AUV in the "right place at the right time"
One application of this effort is to develop derive the basic principles behind model synthesis for adaptive robot sampling for dynamic ocean features through the implementation of an end-to-end autonomous prediction and tasking system. A second application is to increase the utility of minimally actuated vehicles, e.g., gliders and floats.

Funding Source(s): Office of Naval Research, National Science Foundation, QUT ECARD Program


Geometric Control Theory

The general submerged rigid body belongs to a class of simple mechanical control systems whose Lagrangian is of the form kinetic energy minus potential energy. There are many formulations for modelling such systems, and we choose a differential geometric formulation. The configuration space for an AUV corresponds naturally to a differentiable manifold in a one to one manner, and motions can be described via an affine connection control system on this manifold. This project focuses on increasing the autonomy of autonomous underwater vehicles by improving navigation capabilities via extensions to this affine geometric control theory. We are interested in extending the current notion of a kinematic reduction, investigating new formulations of an affine connection, and incorporating a wide range of external disturbances into the differential geometric architecture. Additionally, we are interested in developing a geometric formulation for vehicles that can alter their mass and buoyancy, e.g., gliders.

Funding Source(s): National Science Foundation


Three-Dimensional Reconstructions of Coral Reefs for Assessment and Management

Coral reefs are underwater structures composed of calcium carbonate secreted by vast colonies of tiny living animals called corals. Reefs prosper in warm, shallow, clear, sunny and agitated waters,and are referred to as rainforests of the sea since they form some of the most diverse ecosystems on Earth. Although they occupy less than 0.1% of the oceans, they contain more than 25% of all known marine species. Coral reefs have a significant socio-economic footprint, with their annual value estimated at >USD 375 billion. However, these are fragile ecosystems, that are under enormous threat from climate change, ocean acidification, over-fishing, overuse of reef resources, urban and agricultural runoff and water pollution. Properly assessing the growth or decline of reef environments can indicate the health of the surrounding ecosystem and is important to broadening our knowledge of the effects of global climate change. Computing an accurate estimate of the total surface area, volume and mass of an organism is considered of fundamental and practical importance in benthic ecology, specifically for understanding the complex dynamics of energy flow, cycling of organic matter, and carbonate production in aquatic ecosystems. A specific example for such a necessity lies in the study of coral reef ecology. Here, correlation between biomass, symbiotic dinoflagellate density, chlorophyll concentration and respiration rates can be related to surface area and volumetric measurements. Although physical measurements seem easier to acquire than biological or chemical measurements, prior and existing techniques for estimating the physical parameters of coral colonies are problematic, time-intensive, can require complicated laboratory procedures and need repeated sampling. Few non-intrusive methods exist for in situ estimation, and many methods ultimately resort in destruction of the colony. In this research, we propose a non-intrusive, and repeatable in situ method for generating a 3-D reconstruction of a coral reef environment for the purpose of estimating physical parameters, such as surface area, volume and long-term variability.

Funding Source(s): Google, QUT ECARD Program