Almost two thirds of the Earth's surface is covered in ocean, and yet, only about 5% of it is mapped. There are an unknown amount of sunken ships, planes, and other artifacts hidden below the sea. Extensive search via boat and a sonar tow fish following a standard lawnmower pattern is used to identify sites of interest. Then, if a site has been determined to potentially be historically significant, the most common next step is a survey by either a human dive team or remotely operated vehicle. These are time consuming, error prone, and potentially dangerous options, but autonomous underwater vehicles (AUVs) are a possible solution. This thesis introduces a system for automatically generating paths for AUVs to survey and map shipwrecks. Most AUVs include software to set a lawnmower path for a given region of ocean, and individualized paths can be set via specifying GPS encoded nodes for the AUV to pass through. This thesis presents an algorithm for generating an individualized path that permits the AUV, equipped with a camera to "see" all sides of a region of interest (i.e. a shipwreck). This allows the region of interest to be completely documented. Photogrammetry can then be used to reconstruct a three-dimensional model, but a path is needed to do so. Paths are generated by a probabilistic roadmap algorithm that uses a rapidly-exploring random tree to quickly cover the volume of exploration space and generate small maps with good coverage. The roadmap is constructed out of nodes, each having its own weight. The weight of a given node is calculated using an objective function which measures an approximate view coverage by casting rays from the virtual view and intersecting them with the region of interest. In addition, the weight of a node is increased if this node allows the AUV to see a new side of the region of interest. In each iteration of the algorithm, a node to expand off of is selected based off its location in space or its high weight, a new node with a given amount of freedom is generated, and then added to the roadmap. The algorithm has degrees of freedom in position, pitch, and yaw as well as the objective function to encourage the path to see all sides of the region of interest. Once all sides of the region of interest have been viewed, a path is determined to be complete. The algorithm was tested in a virtual world where the virtual camera acted as the AUV. All of the images collected from our automatically generated path were used to create 3D models and point clouds using photogrammetry. To measure the effectiveness of our paths versus the pre-packaged lawnmower paths, the 3D models and point clouds created from our algorithm were compared to those generated from running a standard lawnmower pattern. The paths generated by our algorithm captured images that could be used in a 3D reconstruction which were more detailed and showed better coverage of the region of interest than those from the lawnmower pattern.

<p>Many of the applications that have been proposed for future small unmanned aerial vehicles (UAVs) are at low altitude in areas with many obstacles. A vital component for successful navigation in such environments is a path planner that can find collision free paths for the UAV.</p><p>Two popular path planning algorithms are the probabilistic roadmap algorithm (PRM) and the rapidly-exploring random tree algorithm (RRT).</p><p>Adaptations of these algorithms to an unmanned autonomous helicopter are presented in this thesis, together with a number of extensions for handling constraints at different stages of the planning process.</p><p>The result of this work is twofold:</p><p>First, the described planners and extensions have been implemented and integrated into the software architecture of a UAV. A number of flight tests with these algorithms have been performed on a physical helicopter and the results from some of them are presented in this thesis.</p><p>Second, an empirical study has been conducted, comparing the performance of the different algorithms and extensions in this planning domain. It is shown that with the environment known in advance, the PRM algorithm generally performs better than the RRT algorithm due to its precompiled roadmaps, but that the latter is also usable as long as the environment is not too complex. The study also shows that simple geometric constraints can be added in the runtime phase of the PRM algorithm, without a big impact on performance. It is also shown that postponing the motion constraints to the runtime phase can improve the performance of the planner in some cases.</p> / Report code: LiU–Tek–Lic–2006:10.

Many of the applications that have been proposed for future small unmanned aerial vehicles (UAVs) are at low altitude in areas with many obstacles. A vital component for successful navigation in such environments is a path planner that can find collision free paths for the UAV. Two popular path planning algorithms are the probabilistic roadmap algorithm (PRM) and the rapidly-exploring random tree algorithm (RRT). Adaptations of these algorithms to an unmanned autonomous helicopter are presented in this thesis, together with a number of extensions for handling constraints at different stages of the planning process. The result of this work is twofold: First, the described planners and extensions have been implemented and integrated into the software architecture of a UAV. A number of flight tests with these algorithms have been performed on a physical helicopter and the results from some of them are presented in this thesis. Second, an empirical study has been conducted, comparing the performance of the different algorithms and extensions in this planning domain. It is shown that with the environment known in advance, the PRM algorithm generally performs better than the RRT algorithm due to its precompiled roadmaps, but that the latter is also usable as long as the environment is not too complex. The study also shows that simple geometric constraints can be added in the runtime phase of the PRM algorithm, without a big impact on performance. It is also shown that postponing the motion constraints to the runtime phase can improve the performance of the planner in some cases. / <p>Report code: LiU–Tek–Lic–2006:10.</p>

The motion planning problem can be formulated as a Markov decision process (MDP), if the uncertainties in the robot motion and environments can be modeled probabilistically. The complexity of solving these MDPs grow exponentially as the dimension of the problem increases and hence, it is nearly impossible to solve the problem even without constraints. Using hierarchical methods, these MDPs can be transformed into a semi-Markov decision process (SMDP) which only needs to be solved at certain landmark states. In the deterministic robotics motion planning community, sampling based algorithms like probabilistic roadmaps (PRM) and rapidly exploring random trees (RRTs) have been successful in solving very high dimensional deterministic problem. However they are not robust to system with uncertainties in the system dynamics and hence, one of the primary objective of this work is to generalize PRM/RRT to solve motion planning with uncertainty. We first present generalizations of randomized sampling based algorithms PRM and RRT, to incorporate the process uncertainty, and obstacle location uncertainty, termed as "generalized PRM" (GPRM) and "generalized RRT" (GRRT). The controllers used at the lower level of these planners are feedback controllers which ensure convergence of trajectories while mitigating the effects of process uncertainty. The results indicate that the algorithms solve the motion planning problem for a single agent in continuous state/control spaces in the presence of process uncertainty, and constraints such as obstacles and other state/input constraints. Secondly, a novel adaptive sampling technique, termed as "adaptive GPRM" (AGPRM), is proposed for these generalized planners to increase the efficiency and overall success probability of these planners. It was implemented on high-dimensional robot n-link manipulators, with up to 8 links, i.e. in a 16-dimensional state-space. The results demonstrate the ability of the proposed algorithm to handle the motion planning problem for highly non-linear systems in very high-dimensional state space. Finally, a solution methodology, termed the "multi-agent AGPRM" (MAGPRM), is proposed to solve the multi-agent motion planning problem under uncertainty. The technique uses a existing solution technique to the multiple traveling salesman problem (MTSP) in conjunction with GPRM. For real-time implementation, an ?inter-agent collision detection and avoidance? module was designed which ensures that no two agents collide at any time-step. Algorithm was tested on teams of homogeneous and heterogeneous agents in cluttered obstacle space and the algorithm demonstrate the ability to handle such problems in continuous state/control spaces in presence of process uncertainty.