This thesis is to develop a novel polishing tool system fitted for convex and concave symmetrical surface of combined surface. There are two design goals in this system. First, this system can be used to polish a concave or convex cone surface with various dimensions and angle cones by adjusting its geometric feature of structure. Second, this polishing tool is expected to have high life expectancy in real applications. Because of the advantages, an efficiency and controllable polishing system would be developed.An inference process, based on a top-down planning strategy, was used to obtain the concept design of polishing tool. There are two major parts in the structure of polishing tool system. The first one is its elastic structure. Both its geometric configuration and loading applied at work are adjustable. The second one is the polishing tool of cylindrical shape. With this specific geometric feature, the effect of tool wear on polishing rate is minimized. The finite element method was adopted to analyze the deformation characteristics of the elastic structure. Accordingly, an optimal design about the shape and dimension of elastic structure was determined. The experimental study showed that the developed polishing system had the property of high repeatability in machining rate. It was also demonstrated that the machining rate of system was insensitive to the tool wear during the polishing process. This advantage may allow this system to gain significant benefit in reducing the need of tool replacement. Finally, it was shown that the experimental trends of machining rate due to the change of applied load or polishing speed followed that of cylindrical polishing system. Both of them can be properly predicted based on the lubricating theories.

One of the significant weaknesses of the design speed concept is that it uses the design speed of the most restrictive geometric element as the design speed of the entire road. This leads to potential inconsistencies among successive sections of a road. Previous studies documented that a uniform design speed does not guarantee consistency on rural two-lane facilities. It is therefore reasonable to assume that similar inconsistencies could be found on rural four-lane non-freeway highways. The operating speed-based method is popularly used in other countries for examining design consistency. Numerous studies have been completed on rural two-lane highways for predicting operating speeds. However, little is known for rural four-lane non-freeway highways. This study aims to develop operating speed prediction models for horizontal curves on rural four-lane non-freeway highways using 74 horizontal curves. The data analysis showed that the operating speeds in each direction of travel had no statistical differences. However, the operating speeds on inside and outside lanes were significantly different. On each of the two lanes, the operating speeds at the beginning, middle, and ending points of the curve were statistically the same. The relationships between operating speed and design speed for inside and outside lanes were different. For the inside lane, the operating speed was statistically equal to the design speed. By contrary, for the outside lane, the operating speed was significantly lower than the design speed. However, the relationships between operating speed and posted speed limit for both inside and outside lanes were similar. It was found that the operating speed was higher than the posted speed limit. Two models were developed for predicting operating speed, since the operating speeds on inside and outside lanes were different. For the inside lane, the significant factors are: shoulder type, median type, pavement type, approaching section grade, and curve length. For the outside lane, the factors included shoulder type, median type, approaching section grade, curve length, curve radius and presence of approaching curve. These factors indicate that the curve itself does mainly influence the drivers speed choice.

As robotic systems move from well-controlled settings to increasingly unstructured environments, they are required to operate in highly dynamic and cluttered scenarios. Finding an object, estimating its pose, and tracking its pose over time within such scenarios are challenging problems. Although various approaches have been developed to tackle these problems, the scope of objects addressed and the robustness of solutions remain limited. In this thesis, we target a robust object perception using visual sensory information, which spans from the traditional monocular camera to the more recently emerged RGB-D sensor, in unstructured environments. Toward this goal, we address four critical challenges to robust 6-DOF object pose estimation and tracking that current state-of-the-art approaches have, as yet, failed to solve. The first challenge is how to increase the scope of objects by allowing visual perception to handle both textured and textureless objects. A large number of 3D object models are widely available in online object model databases, and these object models provide significant prior information including geometric shapes and photometric appearances. We note that using both geometric and photometric attributes available from these models enables us to handle both textured and textureless objects. This thesis presents our efforts to broaden the spectrum of objects to be handled by combining geometric and photometric features. The second challenge is how to dependably estimate and track the pose of an object despite the clutter in backgrounds. Difficulties in object perception rise with the degree of clutter. Background clutter is likely to lead to false measurements, and false measurements tend to result in inaccurate pose estimates. To tackle significant clutter in backgrounds, we present two multiple pose hypotheses frameworks: a particle filtering framework for tracking and a voting framework for pose estimation. Handling of object discontinuities during tracking, such as severe occlusions, disappearances, and blurring, presents another important challenge. In an ideal scenario, a tracked object is visible throughout the entirety of tracking. However, when an object happens to be occluded by other objects or disappears due to the motions of the object or the camera, difficulties ensue. Because the continuous tracking of an object is critical to robotic manipulation, we propose to devise a method to measure tracking quality and to re-initialize tracking as necessary. The final challenge we address is performing these tasks within real-time constraints. Our particle filtering and voting frameworks, while time-consuming, are composed of repetitive, simple and independent computations. Inspired by that observation, we propose to run massively parallelized frameworks on a GPU for those robotic perception tasks which must operate within strict time constraints.