Kinematic synthesis of adjustable robotic mechanisms

Conventional hard automation, such as a linkage-based or a cam-driven system, provides high speed capability, and repeatability but not the flexibility required in many industrial applications. The conventional mechanisms, that are typically single-degree-of-freedom systems, are being increasingly replaced by multi-degree-of-freedom multi- actuators driven by logic controllers. Although this new trend in sophistication provides greatly enhanced flexibility, there are many instances where the flexibility needs are exaggerated and the associated complexity is unnecessary. Traditional mechanism-based hard automation, on the other hand, can neither fulfill multi-task requirements nor are cost-effective mainly due to lack of methods and tools to design-in flexibility. This dissertation attempts to bridge this technological gap by developing Adjustable Robotic Mechanisms (ARMs) or "programmable mechanisms" as a middle ground between high speed hard automation and expensive serial jointed-arm robots.;This research introduces the concept of adjustable robotic mechanisms towards cost-effective manufacturing automation. A generalized analytical synthesis technique has been developed to support the computational design of ARMs that lays the theoretical foundation for synthesis of adjustable mechanisms. The synthesis method developed in this dissertation, called generalized adjustable dyad and triad synthesis, advances the well-known Burmester theory in kinematics to a new level. While this method provides planar solutions, a novel patented scheme is utilized for converting prescribed three-dimensional motion specifications into sets of planar projections. This provides an analytical and a computational tool for designing adjustable mechanisms that satisfy multiple sets of three-dimensional motion specifications. Several design issues were addressed; including adjustable parameter identification, branching defect, and mechanical errors. An efficient mathematical scheme for identification of adjustable member was also developed. The analytical synthesis techniques developed in this dissertation were successfully implemented in a graphic-intensive user-friendly computer program. A physical prototype of a general purpose adjustable robotic mechanism has been constructed to serve as a proof-of-concept model.