Júnior Anderson Rodrigues da Silva's dissertation, "Piecewise linear continuous-curvature path planning for autonomous vehicles," presents a method for planning continuous-curvature paths for autonomous vehicles. The work focuses on modeling the road network using GPS data and planning paths that consider the vehicle's physical constraints and comfort level. The approach uses clothoids, circular arcs, and straight lines to create smooth paths that ensure the vehicle can navigate urban environments safely and comfortably. The road network is modeled by defining lanes, intersections, and roundabouts based on GPS trajectories. The path planning process involves computing a global route, then generating intersection, roundabout, and lane change paths that adhere to traffic rules and vehicle capabilities. The dissertation also includes a detailed analysis of the kinematic model of car-like robots, the configuration space, and the computation of discrete curvature. The work is validated through simulations and real-world scenarios, demonstrating the effectiveness of the proposed path planning approach in urban environments. The key contributions include a sparse road network model, real-time lane change and roundabout path planners, and a continuous-curvature path planning method that ensures comfort and feasibility for autonomous vehicles. The dissertation is structured into several chapters covering the technical background, related work, path planning for intersections, lane changes, and roundabouts, global path planning, and results. The work is supported by a detailed list of figures, tables, and references, and it concludes with important remarks and future work directions.Júnior Anderson Rodrigues da Silva's dissertation, "Piecewise linear continuous-curvature path planning for autonomous vehicles," presents a method for planning continuous-curvature paths for autonomous vehicles. The work focuses on modeling the road network using GPS data and planning paths that consider the vehicle's physical constraints and comfort level. The approach uses clothoids, circular arcs, and straight lines to create smooth paths that ensure the vehicle can navigate urban environments safely and comfortably. The road network is modeled by defining lanes, intersections, and roundabouts based on GPS trajectories. The path planning process involves computing a global route, then generating intersection, roundabout, and lane change paths that adhere to traffic rules and vehicle capabilities. The dissertation also includes a detailed analysis of the kinematic model of car-like robots, the configuration space, and the computation of discrete curvature. The work is validated through simulations and real-world scenarios, demonstrating the effectiveness of the proposed path planning approach in urban environments. The key contributions include a sparse road network model, real-time lane change and roundabout path planners, and a continuous-curvature path planning method that ensures comfort and feasibility for autonomous vehicles. The dissertation is structured into several chapters covering the technical background, related work, path planning for intersections, lane changes, and roundabouts, global path planning, and results. The work is supported by a detailed list of figures, tables, and references, and it concludes with important remarks and future work directions.