No description available.

No description available.

Perceiving the surrounding multiple vehicles robustly and effectively is a very important step in building Advanced Driving Assistant System (ADAS) or autonomous vehicles. This thesis presents the design of the Light Detection and Ranging (LiDAR) perception system which consists of several sub-tasks: ground detection, object detection, object classification, and object tracking. It is believed that accomplishing these sub-tasks will provide a guideline to a vast range of potential autonomous vehicles applications. More specifically, a probability occupancy grid map based approach was developed for ground detection to address the issues of over-segmentation, under-segmentation and slow-segmentation by non-flat surface. Given the non-ground points, point cloud clustering algorithm is developed for object detection by using a Radially Bounded Nearest Neighbor (RBNN) method on the static Kd-tree. To identify the object, a supervised learning approach based on our LiDAR sensor for vehicle type classification is proposed. The proposed classification algorithm is used to classify the object into four different types: ``Sedan'', ``SUV'', ``Van'', and ``Truck''. To handle disturbances and motion uncertainties, a generalized form of Smooth Variable Structure Filter (SVSF) integrated with a combination of Hungarian algorithm (HA) and Probability Data Association Filter (PDAF), referred to as GSVSF-HA/PDAF, is developed. The developed approach is to overcome the multiple targets data association in the content of dynamics environment where the distribution of data is unpredictable. Last but not the least, a comprehensive experimental evaluation for each sub-task is presented to validate the robustness and effectiveness of our developed perception system. / Thesis / Doctor of Philosophy (PhD)

The introduction of electric autonomous Unmanned Arial Vehicles (UAVs) in cities is considered the ultimate disruptive sustainable technological solution due to the promised speed, affordability, and significant greenhouse gas (GHG) emission reductions. The integration of UAVs into the future smart city fabric offers a wide range of applications. In particular, UAVs are ideal for last-mile operation, which is expected to reduce delivery costs, GHG emissions, and delivery time compared to light trucks and other traditional delivery methods. As UAVs operate in the city airspace, and with the current generation of older cities, several technological challenges arise with the anticipated proliferation of heterogeneous UAV fleets in low-altitude airspace of dense urban areas. Being a fairly new disruptive technology with no real-world operation data, the literature only considers a few of the system design parameters and often disregards the impact of other essential parameters such as Kinematics and airspace policies. This leads to significant uncertainty in the estimated UAV energy consumption, ranges, and emissions yielding inaccurate conclusions regarding the full system design predilections. Therefore, an effective UAV system design should strive to understand the broad spectrum of parameters’ impacts to optimize the integration and operation. Towards that end, this research aims at investigating the different UAV system design parameters and their intertwined impacts on operation efficiency to obtain accurate system optimization results. The research utilized several datasets for the delivery demand and digital-twin city model data of Toronto, Ontario, Canada. The research employed a state-of-the-art flexible energy use model for UAVs calibrated to experimental measurements to generate a minimum-energy trajectory along with several proposed novel airspace discretization, trajectory optimization, and charging infrastructure allocation optimization models. In this respect, this dissertation quantified the impact of airspace policies, discretization, and trajectory generation on the energy consumption of UAVs. Furthermore, it unveiled the operation uncertainties and their implications on the cost, emissions, and allocated charging infrastructure demand. Unlike the UAV literature, our research included all system design parameters and their impact on the performance metrics. The dissertation also proposes a novel combined airspace discretization and trajectory generation algorithm for optimal UAV energy consumption, airspace capacity maximization, airspace traffic control, and off-grid solar charging station allocation. For instance, it is found that UAV deployment with carefully tailored airspace policies in delivery could reduce GHG emissions in the freight sector by up to 35% compared to EVs. Furthermore, the research highlighted how building integrated photovoltaic BIPV upgrades with associated buildings can eliminate GHG emissions and significantly reduce the decarbonization price through associated savings and excess generated electricity. Overall, this research presents a unique contribution to the knowledge of UAV research for practitioners, policymakers, and academia. / Thesis / Doctor of Philosophy (PhD)

No / This chapter describes a test-generation approach that takes as input a model S of the expected behavior of a robotic system and seeds faults into S, leading to a set of mutants of S. Given a mutant M of S, we check whether M is a valid implementation of S, and, if it is not, we find a test case that demonstrates this: a test case that reveals the seeded fault. In order to automate this approach, we used the Wodel tool to seed faults and a combination of two tools, RoboTool and FDR, to generate tests that detect the seeded faults. The result is an overall test-generation technique that can be automated and that derives test cases that are guaranteed to find certain faults. / EPSRC grant EP/R025134/2 RoboTest: Systematic Model-Based Testing and Simulation of Mobile Autonomous Robots, Spanish MINECO-FEDER grant FAME RTI2018-093608-B-C31 and the Comunidad de Madrid project FORTE-CM S2018/TCS-4314.

Multi-robot systems (MRS) offer many performance benefits over single robots for tasks that can be completed by one robot. They offer potential redundancies to the system to improve robustness and allow tasks to be completed in parallel. These benefits, however, can be quickly offset by losses in productivity from diminishing returns caused by interference between robots and communication problems. This dissertation developed and evaluated MRS control architectures to solve the dynamic multi-robot autonomous routing problem. Dynamic multi-robot autonomous routing requires robots to complete a trip from their initial location at the time of task allocation to an assigned destination. The primary concern for the control architectures was how well the communication requirements and overall system performance scaled as the number of robots in the MRS got larger. The primary metrics for evaluation of the controller were the effective robot usage rate and the bandwidth usage. This dissertation evaluated several different approaches to solving dynamic multi-robot autonomous routing. The first three methods were based off of common MRS coordination approaches from previous research. These three control architectures with distributed control without communication (a swarm-like method), distributed control with communication, and centralized control. An additional architecture was developed to solve the problem in a way that scales better as the number of robots increase. This architecture, mixed mode autonomy, combines the strengths of distributed control with communication and centralized control. Like distributed control with communication, mixed mode autonomy's performance degrades gracefully with communication failures and is not dependent on a single controller. Like centralized control, there is oversight from a central controller to ensure repeatable high performance of the system. Each of the controllers other than distributed control without communication is based on building world models to facilitate coordination of the routes. A second variant of mixed mode autonomy was developed to allow robots to share parts of their world models with their peers when their models were incomplete or outdated. The system performance was evaluated for three example applications that represent different cases of dynamic multi-robot autonomous routing. These example applications were the automation of open pit mines, container terminals, and warehouses. The effective robot usage rates for mixed mode autonomy were generally significantly higher than the other controllers with a higher numbers of robots. The bandwidth usage was also much lower. These performance trends were also observed across a wide range of operating conditions for dynamic multi-robot autonomous routing. The original contributions from this work were the development of a new MRS control architecture, development of system model for the dynamic multi-robot autonomous routing problem, and identification of the tradeoffs for MRS design for the dynamic multi-robot autonomous routing problem. / Ph. D.

Autonomous robots are expected to interact with their dynamic changing environment. This interactions requires certain level of behavior based Intelligence, which facilitates the dynamic adaptation of the robot behavior accordingly with his surrounding environment. Many researches have been done in biological information processing systems to model the behavior of an autonomous robot. The Artificial Immune System (AIS) provides new paradigm suitable for dynamic problem dealing with unknown environment rather than a static problem. The immune system has some features such as memory, tolerance, diversity and more features that can be used in engineering applications. The immune system has an important feature called meta-dynamics in which new species of antibodies are produced continuously from the bone marrow. If the B-Cell (robot) cannot deal with the current situation, new behaviors (antibodies) should be generated by the meta dynamics function. This behavior should be incorporated into the existing immune system to gain immunity against new environmental changes. We decided to use a feed forward Artificial Neural Network (ANN) to simulate this problem, and to build the AIS memory. Many researchers have tried to tackle different points in mimicking the biological immune system, but no one previously has proposed such an acquired memory. This contribution is made as a "proof of concept" to the field of biological immune system simulation as a start of further research efforts in this direction. Many applications can potentially use our designed Neuro-Immune Network (NIN), especially in the area of autonomous robotics. We demonstrated the use of the designed NIN to control a robot arm in an unknown environment. As the system encounters new cases, it will increase its ability to deal with old and new situations encountered. This novel technique can be applied to many robotics applications in industry, where autonomous robots are required to have adaptive behavior in response to their environmental changes. Regarding future work, the use of VLSI neural networks to enhance the speed of the system for real time applications can be investigated along with possible methods of design and implementation of a similar VLSI chip for the AIN. / Ph. D.

This research study presents two control solutions, PID and the novel hyper-clothoid control strategy, to autonomously navigate a car. These waypoint navigation solutions smoothly connect the given waypoints with C1 continuity using Hermite cubic splines which is used as a reference path for the controller to track. The PID controller uses lateral and heading error to generate a steering profile for the vehicle to track the constructed reference path. A novel real time solution is presented as the second control strategy which involves constructing clothoids to generate a steering profile. The resultant car trajectory preserves curvature and curvature rate continuity. A simulation test bench was developed in MATLAB and Simulink. Six benchmark waypoint datasets have been used for regression testing and validating the algorithms. Both the proposed control strategies have been implemented on a 2017 GM Chevy Bolt EV. A real time operating system (QNX) has been used and was time-synced with the localization suite in the test vehicle. Closed loop results with accuracies up to 50 cm of lateral error have been achieved using the test vehicle. / Doctor of Philosophy / The research into self-driving cars has been one of the most sought out areas these past couple of decades. There are many components into building a self-driving car - Sensing, Perception, Localization, Navigation. Lot of strategies have been developed over the years with waypoint navigation being the most widely used for navigation an autonomous vehicle. Waypoint Navigation utilizes GPS data to move the car from one point to the other. The traditional process of this strategy involves two parts - curve fitting between waypoints and using a control scheme to track the path with the car. Numerous methods have been developed to fit a curve in between two points. Most of these methods use a variant of 3rd degree or higher order polynomials . Also different control strategies have been developed to track the generated path. Model predictive control strategies are among the popular control architectures used for this purpose. This work proposes a novel method to track a path using clothoids. The proposed algorithm has a novel approach of integrating the path construction and control strategy. The algorithm also has a low computational requirement making it highly suitable for implementation in real-time.

In the context of path planning, the non-signalized intersections are always a challenging scenario due to the mixture of traffic flow. Most path planning algorithms use the information at the current time instance to generate an optimal path. Because of the dynamics of the non-signalized intersections, iteratively generating a path in a high frequency is necessary, resulting in an enormous waste of computational resources. Rapidly-exploring Random Tree (RRT) as an effective local path planning methodology can determine a feasible path in the static environment. Few improvements are proposed to adopt the RRT to the non-signalized intersections. Gaussian Processes Regression (GPR) is used to predict the other vehicles' future location. The location information in the current and future time instance is used to generate a probability position map. The map not only avoids useless sampling procedures but also increases the speed of generating a path. The optimal steering strategy is deployed to guarantee the trajectory is collision-free in both current and future time frames. Overall, the proposed probabilistic RRT algorithm can select a collision-free path in the non-signalized intersections by combining the GPR, probability position map, and optimal-steering. / Master of Science / Path planning problem is a challenge in the non-signalized intersections. Many path planning algorithms can generate an optimal path in the space domain but not in the time domain. Thus, the algorithms need to run iteratively at a high frequency to ensure the path's optimality in the time domain. By combining prediction and the standard RRT path planning algorithm, the resulting path ensures to be optimal in the space and time domain.

Autonomous vehicle development is a rapidly growing field that has vast possibilities for both military and commercial applications. Removing people from dangerous tasks will save lives. Continued research is necessary in order to build these new technologies and mature those already established. One area of potential in the unmanned vehicle community is that of fully autonomous cooperation. This area of research will allow multiple unmanned platforms to perform new functions on a larger scale by combining their capabilities in a coordinated manner. This thesis addresses the emerging need of research related to fully autonomous cooperation between a heterogeneous team of vehicles, by taking a system level approach and integrating the necessary technologies. Software was developed and then tested that combines an unmanned ground vehicle and an unmanned aerial vehicle in order to perform a task that utilizes the strengths of each platform. The ground vehicle is programmed with a route for which it sends look-ahead waypoints to the aircraft. As it traverses the route, the aircraft searches for possible targets. If a target is detected, the approximate coordinates are sent over the network and the ground vehicle then further localizes and inspects the target. Once the inspection is completed, the ground vehicle continues on its previous route. This thesis demonstrates that pairing ground and aerial vehicles in a fully autonomous target localization problem can indeed provide a team functioning more efficiently than either alone. / Master of Science