Many robot object interactions require that an object is firmly held, and that the grasp remains stable during the whole manipulation process. Based on grasp wrench space, this thesis address the problems of measuring the grasp sensitivity against friction changes, planning contacts and hand configurations on mesh and point cloud representations of arbitrary objects, planning adaptable grasps and finger gaiting for keeping a grasp stable under various external disturbances, as well as learning of grasping manifolds for more accurate reachability and inverse kinematics computation for multifingered grasping. 

Firstly, we propose a new concept called friction sensitivity, which measures how susceptible a specific grasp is to changes in the underlying frictionc oefficients. We develop algorithms for the synthesis of stable grasps with low friction sensitivity and for the synthesis of stable grasps in the case of small friction coefficients.  

Secondly, for fast planning of contacts and hand configurations for dexterous grasping, as well as keeping the stability of a grasp during execution, we present a unified framework for grasp planning and in-hand grasp adaptation using visual, tactile and proprioceptive feedback. The main objective of the proposed framework is to enable fingertip grasping by addressing problems of changed weight of the object, slippage and external disturbances. For this purpose, we introduce the Hierarchical Fingertip Space (HFTS) as a representation enabling optimization for both efficient grasp synthesis and online finger gaiting. Grasp synthesis is followed by a grasp adaptation step that consists of both grasp force adaptation through impedance control and regrasping/finger gaiting when the former is not sufficient. 

Lastly, to improve the efficiency and accuracy of dexterous grasping and in-hand manipulation, we present a system for fingertip grasp planning that incrementally learns a heuristic for hand reachability and multi-fingered inverse kinematics. During execution the system plans and executes fingertip grasps using Canny’s grasp quality metric and a learned random forest based hand reachability heuristic. In the offline module, this heuristic is improved based on a grasping manifold that is incrementally learned from the experiences collected during execution.

This thesis presents planning algorithms for three different robot manipulation tasks: fingertip grasping, object placing and rearranging. Herein, we place special attention on addressing these tasks in the presence of obstacles. Obstacles are frequently encountered in human-centered environments and constrain a robot's motion and ability to manipulate objects. In narrow shelves, for example, even the common task of pick-and-place becomes challenging. A shelf is difficult to navigate and many potential grasps and placements are inaccessible. Hence, to solve such tasks, specialized manipulation planning algorithms are required that can cope with the presence of obstacles.

For fingertip grasping, we first present a framework to learn models that encode which grasps a given dexterous robot hand can reach. These models are then used to facilitate planning and optimization of fingertip grasps. Next, we address the presence of obstacles and integrate fingertip grasp and motion planning to generate grasps that are reachable by a robot in complex scenes.

For object placing, we first present an algorithm that plans the placement of a grasped object among obstacles so that a user-given placement objective is maximized. We then extend this algorithm, and incorporate planning in-hand manipulation to increase the set of placements a robot can reach.

Lastly, we go beyond pure collision avoidance and study object rearrangement planning. Specifically, we consider the special case of non-prehensile rearrangement, where a robot rearranges multiple objects through pushing. First, we present how a kinodynamic motion planning algorithm can be augmented with learned models to rearrange a few target objects among movable and static obstacles. We then present how we can use Monte Carlo tree search to solve a large-scale rearrangement problem, where a robot is tasked to spatially sort many objects according to a user-assigned class membership.

Den här avhandlingen presenterar planeringsalgoritmer för tre olika ma-nipulationsuppgifter för robotar i närheten av hinder: att greppa med hjälpav fingertopparna, att placera objekt och att arrangera om flertalet objekt iolika konfigurationer. Hinder finns oftast i människors miljöer och begränsaren robots rörelse och förmåga att manipulera objekt. Till exempel är den tillsynes enkla uppgiften att hämta och placera objekt i smala hyllor mycketsvår för robotar. Planering av manipulering i detta fall blir svårt därför attmånga rörelser och grepp kommer att kollidera med hinder. För att lösa dessauppgifter behöver man speciella algoritmer som kan hantera dessa hinder imiljön.

För grepp med fingertoppar presenterar vi ett ramverk för inlärning avmodeller som representerar vilka grepp en robothand kan utföra. Dessa mo-deller används sedan för att planera och optimera grepp som sker med hjälpav fingertopparna. Därefter integreras planeringsalgoritmen med rörelsepla-nering för att kunna planera fingertoppsgrepp där också hinder existerar iobjektets närhet.

För objektsplaceringar presenterar vi ett ramverk för att planera hur enrobot kan transportera ett greppat objekt till en placering som optimerar engiven kostnadsfunktion. Därefter utvecklar vi detta ramverk för att visa hurså kallad in-hand manipulation (att byta grepp utan att släppa objektet) kanöka antalet möjliga placeringar som roboten kan utföra.

Till slut utökar vi våra metoder bortom endast undvikande av kollisionoch studerar planering för omarrangering av objekt. Vi studerar det speciellafallet non-prehensile rearrangement där en robot måste ordna om flertaletobjekt genom att skjuta dem framför sig. Vi presenterar först hur en kino-dynamisk rörelseplaneringsalgoritm kan förbättras genom inlärda modeller.Detta görs i syftet att hitta rörelsesekvenser som kan ordna flertalet objekt tillgivna konfigurationer. Sedan presenterar vi hur vi kan använda Monte Carlo-trädsök för att lösa ett stort sådant problem där objekt ska om-arrangeras. Idetta problem ska roboten sortera objekt enligt kategorier som en användarehar specifierat.