<p>Multi-robot systems of autonomous mobile robots offer many benefits but also many challenges. This work addresses collision avoidance of robots solving continuous problems in known environments. The approach to handling collision avoidance is here to enhance a motion planning method for single-robot systems to account for auxiliary robots. A few assumptions are made to put the focus of the work on path planning, rather than on localization.</p><p>A method, based on exact cell decomposition and extended with a few rules, was developed and its consistency was proven. The method is divided into two steps: path planning, which is off-line, and path monitoring, which is on-line. This work also introduces the notion of<em>path obstacle</em>, an essential tool for this kind of path planning with many robots.</p><p>Furthermore, an implementation was performed on a system of omni-directional robots and tested in simulations and experiments. The implementation practices centralized control, by letting an additional computer handle the motion planning, to relieve the robots of strenuous computations.</p><p>A few drawbacks with the method are stressed, and the characteristics of problems that the method is suitable for are presented.</p> / QC 20100705

Inspired by the Rapidly-exploring Random Tree (RRT) data-structure and algorithm for path planning, we introduce an approach for spanning physical space with a group of simple mobile robots. Emphasizing minimalism and using only InfraRed and contact sensors for communication, our position unaware robots physically embody elements of the tree. Although robots are fundamentally constrained in the spatial operations they may perform, we show that the approach -implemented on physical robots- remains consistent with the original data-structure idea. In particular, we show that a generalized form of Voronoi bias is present in the construction of the tree, and that such trees have an approximate space-filling property. We present an analysis of the physical system via sets of coupled stochastic equations: the first being the rate-equation for the transitions made by the robot controllers, and the second to capture the spatial process describing tree formation. We also introduce a class of fixed edge length RRTs called lRRT and show that lRRT s have similar space-filling properties to that of RRTs. We are able to provide an understanding of the control parameters in terms of a process mixing-time and show the dependence of the Voronoi bias on an interference parameter which grows as O*sqrt(N).

This thesis investigates cooperative and intelligent control of autonomous multi-robot systems in a dynamic, unstructured and unknown environment and makes significant original contributions with regard to self-deterministic learning for robot cooperation, evolutionary optimization of robotic actions, improvement of system robustness, vision-based object tracking, and real-time performance. A distributed multi-robot architecture is developed which will facilitate operation of a cooperative multi-robot system in a dynamic and unknown environment in a self-improving, robust, and real-time manner. It is a fully distributed and hierarchical architecture with three levels. By combining several popular AI, soft computing, and control techniques such as learning, planning, reactive paradigm, optimization, and hybrid control, the developed architecture is expected to facilitate effective autonomous operation of cooperative multi-robot systems in a dynamically changing, unknown, and unstructured environment. A machine learning technique is incorporated into the developed multi-robot system for self-deterministic and self-improving cooperation and coping with uncertainties in the environment. A modified Q-learning algorithm termed Sequential Q-learning with Kalman Filtering (SQKF) is developed in the thesis, which can provide fast multi-robot learning. By arranging the robots to learn according to a predefined sequence, modeling the effect of the actions of other robots in the work environment as Gaussian white noise and estimating this noise online with a Kalman filter, the SQKF algorithm seeks to solve several key problems in multi-robot learning. As a part of low-level sensing and control in the proposed multi-robot architecture, a fast computer vision algorithm for color-blob tracking is developed to track multiple moving objects in the environment. By removing the brightness and saturation information in an image and filtering unrelated information based on statistical features and domain knowledge, the algorithm solves the problems of uneven illumination in the environment and improves real-time performance.

This thesis investigates cooperative and intelligent control of autonomous multi-robot systems in a dynamic, unstructured and unknown environment and makes significant original contributions with regard to self-deterministic learning for robot cooperation, evolutionary optimization of robotic actions, improvement of system robustness, vision-based object tracking, and real-time performance. A distributed multi-robot architecture is developed which will facilitate operation of a cooperative multi-robot system in a dynamic and unknown environment in a self-improving, robust, and real-time manner. It is a fully distributed and hierarchical architecture with three levels. By combining several popular AI, soft computing, and control techniques such as learning, planning, reactive paradigm, optimization, and hybrid control, the developed architecture is expected to facilitate effective autonomous operation of cooperative multi-robot systems in a dynamically changing, unknown, and unstructured environment. A machine learning technique is incorporated into the developed multi-robot system for self-deterministic and self-improving cooperation and coping with uncertainties in the environment. A modified Q-learning algorithm termed Sequential Q-learning with Kalman Filtering (SQKF) is developed in the thesis, which can provide fast multi-robot learning. By arranging the robots to learn according to a predefined sequence, modeling the effect of the actions of other robots in the work environment as Gaussian white noise and estimating this noise online with a Kalman filter, the SQKF algorithm seeks to solve several key problems in multi-robot learning. As a part of low-level sensing and control in the proposed multi-robot architecture, a fast computer vision algorithm for color-blob tracking is developed to track multiple moving objects in the environment. By removing the brightness and saturation information in an image and filtering unrelated information based on statistical features and domain knowledge, the algorithm solves the problems of uneven illumination in the environment and improves real-time performance.

This thesis investigates cooperative and intelligent control of autonomous multi-robot systems in a dynamic, unstructured and unknown environment and makes significant original contributions with regard to self-deterministic learning for robot cooperation, evolutionary optimization of robotic actions, improvement of system robustness, vision-based object tracking, and real-time performance. A distributed multi-robot architecture is developed which will facilitate operation of a cooperative multi-robot system in a dynamic and unknown environment in a self-improving, robust, and real-time manner. It is a fully distributed and hierarchical architecture with three levels. By combining several popular AI, soft computing, and control techniques such as learning, planning, reactive paradigm, optimization, and hybrid control, the developed architecture is expected to facilitate effective autonomous operation of cooperative multi-robot systems in a dynamically changing, unknown, and unstructured environment. A machine learning technique is incorporated into the developed multi-robot system for self-deterministic and self-improving cooperation and coping with uncertainties in the environment. A modified Q-learning algorithm termed Sequential Q-learning with Kalman Filtering (SQKF) is developed in the thesis, which can provide fast multi-robot learning. By arranging the robots to learn according to a predefined sequence, modeling the effect of the actions of other robots in the work environment as Gaussian white noise and estimating this noise online with a Kalman filter, the SQKF algorithm seeks to solve several key problems in multi-robot learning. As a part of low-level sensing and control in the proposed multi-robot architecture, a fast computer vision algorithm for color-blob tracking is developed to track multiple moving objects in the environment. By removing the brightness and saturation information in an image and filtering unrelated information based on statistical features and domain knowledge, the algorithm solves the problems of uneven illumination in the environment and improves real-time performance. / Applied Science, Faculty of / Mechanical Engineering, Department of / Graduate

Multi-robot systems of autonomous mobile robots offer many benefits but also many challenges. This work addresses collision avoidance of robots solving continuous problems in known environments. The approach to handling collision avoidance is here to enhance a motion planning method for single-robot systems to account for auxiliary robots. A few assumptions are made to put the focus of the work on path planning, rather than on localization. A method, based on exact cell decomposition and extended with a few rules, was developed and its consistency was proven. The method is divided into two steps: path planning, which is off-line, and path monitoring, which is on-line. This work also introduces the notion ofpath obstacle, an essential tool for this kind of path planning with many robots. Furthermore, an implementation was performed on a system of omni-directional robots and tested in simulations and experiments. The implementation practices centralized control, by letting an additional computer handle the motion planning, to relieve the robots of strenuous computations. A few drawbacks with the method are stressed, and the characteristics of problems that the method is suitable for are presented. / QC 20100705

In recent years, we have seen the applications of distributed autonomous mobile robots (DAMRs) in a broad spectrum of areas like search and rescue, disaster management, warehouse, and delivery systems. Although each type of systems employing DAMRs has its specific challenges, they are all limited by energy since the robots are powered by batteries which have not advanced in decades. This motivates the development of energy efficiency for such systems. Although there has been research on optimizing energy for robotic systems, their approaches are from low-level (e.g., mechanic, system control, or avionic) perspectives. They, therefore, are limited to a specific type of robots and not easily adjusted to apply for different types of robots. Moreover, there is a lack of work studying the problem from a software perspective and abstraction. In this thesis, we tackle the problem from a software perspective and are particularly interested in DAMR systems in which a team of networked robots navigating in a physical environment and acting in concert to accomplish a common goal. Also, the primary focus of our work is to design schedules (or plans) for the robots so that they can achieve their goal while spending as little energy as possible. To this end, we study the problem in three different contexts: - Managing reliability and energy consumption tradeoff. That is, we propose that robots verify computational results of one another to increase the corroboration of outputs of our DAMR systems. However, this new feature requires robots to do additional tasks and consume more energy. Thus, we propose approaches to reach a balance between energy consumption and the reliability of results obtained by our DAMR systems. - Extending the operational time of robots. We first propose that our DAMR systems should employ charging stations where robots can come to recharge their batteries. Then, we aim to design schedules for the robots so that they can finish all their tasks while consuming as little energy and time (including the time spent for recharging) as possible. Moreover, we model the working space by a connected (possibly incomplete) graph to make the problem more practical. - Coping with environmental changes. This path planning problem takes into account not only energy limits but also changes in the physical environment, which may result in overheads (i.e., additional time and energy) that robots incur while doing their tasks. To tackle the problem from a software perspective, we first utilize Gaussian Process and Polynomial Regression to model disturbances and energy consumption, respectively, then proposed techniques to generate plans and adjust them when robots detect environmental changes. For each problem, we give a formal description, a transformation to integer (linear) programming, online algorithms, and an online algorithm. Moreover, we also rigorously analyze the proposed techniques by conducting simulations and experiments in a real network of unmanned aerial vehicles (UAVs). / Thesis / Candidate in Philosophy

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.

The objective of this thesis is to develop a strategy that allows robots to safely execute open-loop motion patterns for pre-computed time durations when facing interruptions in communication. By computing the time horizon in which collisions with other robots are impossible, this method allows the robots to move safely despite having no updated information about the environment. As the complexity of multi-robot systems increase, communication failures in the form of packet losses, saturated network channels and hardware failures are inevitable. This thesis is motivated by the need to increase the robustness of operation in the face of such failures. The advantage of this strategy is that it prevents the jerky and unpredictable motion behaviour which often plague robotic systems experiencing communication issues. To compute the safe time horizon, the first step involves constructing reachable sets around the robots to determine the set of all positions that can be reached by the robot in a given amount of time. In order to avoid complications arising from the non-convexity of these reachable sets, analytical expressions for minimum area ellipses enclosing the reachable sets are obtained. By using a fast gradient descent based technique, intersections are computed between a robot’s trajectory and the reachable sets of other robots. This information is then used to compute the safe time horizon for each robot in real time. To this end, provable safety guarantees are formulated to ensure collision avoidance. This strategy has been verified in simulation as well as on a team of two-wheeled differential drive robots on a multi-robot testbed.

Simultaneous localization and mapping (SLAM) is a challenging and vital problem in robotics. It is important in tasks such as disaster response, deep-sea and cave exploration, in which robots must construct a map of an unknown terrain, and at the same time localize themselves within the map. The issue with single- robot SLAM is the relatively high rate of failure in a realistic application, as well as the time and energy cost. In this work, we propose a new approach to decentralized multi-robot SLAM which uses a robot swarm to map the environment. This system is capable of mapping an environment without human assistance and without the need for any additional infrastructure. We assume that 1) no robot possesses sufficient memory to store the entire map of the environment, 2) the communication range of the robots is limited, and 3)there is no infrastructure present in the environment to assist the robot in communicating with others. To cope with these limitations, the swarm system is designed to work as an independent entity. The swarm can deploy new robots towards the region that is yet to be explored, coordinate the communication between the robots by using itself as the communication network and replace any malfunctioning robots. The proposed method proves to be a reliable and robust exploration algorithm. It is shown to be a self-growing mapping network that is able to coordinate among numerous robots and replace any broken robots hence reducing the chance of system failure.