PURPOSE

As a product is assembled in an automated factory, both the product and its individual parts are picked up, reoriented and inserted into subassemblies or fixtures. For a complex product, the number of manipulations could run into the thousands. Parts are picked out of bins and placed into assemblies. Partial assemblies are rotated to allow additional parts to be added. Fasteners are inserted to hold it all together. The machines that perform these tasks are selected for their speed, precision, cost and reliability.

Assembly line designers try to keep the manipulations as simple as possible. Rotations about vertical or horizontal axes are preferred, often of 90° or 180°. These tasks have a well established set of solutions.

However, operations which involve a translation along and a rotation about an axis which is not vertical or horizontal is more challenging to the designer. Additional constraints on the trajec­tory of the object (e.g. obstacle avoidance, or part meshing) increase the difficulties. One solution is to use devices with a high number of degrees of freedom, such as robots. These mechan­ically complex devices perform the tasks, but at penalties in dollars, setup time and maintenance. A second solution is to use a cascading series of simple manipulators. Creating this manipulation pipeline takes a longer design time, and is often more art than science.

This research offers another solution. Moving an object from one place to another doesn’t require six degrees of freedom if the motion of the mechanism is designed with the task in mind.. This research will create low degree-of-freedom machines capable of producing spatial trajectories and rotations. A single “part orienting device” (POD) can be used in an assembly task that might otherwise require a robot or multiple single degree-of-freedom mechanisms.

Spatial devices similar to the proposed PODs exist in research settings, but have yet to be well utilized in a practical environ­ment. The design and manufacturing issues have been too daunting. This research will comprehensively address the design of PODS. (1) A kinematic synthesis method will be derived to enable the designer to create PODS for a prescribed task. (2) An interactive design system will be created to allow the designer to navigate a broad array of choices in a goal-oriented, immersive system. (3) Tools will be created to provide the designer with prototyping and actuating recommendations.

Design of Robot-like Devices for Lean Manufacturing Cells

The goal of this research is to develop a methodology for designing low complexity mechanisms capable of performing a non-trivial translation and reorientation of a part in an automated assembly procedure.  These mechanisms would offer many of the benefits of a robot with lower material costs and programming complexity. 

The design process begins by specifying the desired initial and final positions and orientations of a part.  Further constraints can be placed upon the design, such as error tolerance, maximum acceleration or workspace obstacles.  The design procedure will find the optimal solution based upon the provided constraints and the desire to construct the mechanism economically.

The theoretical impact of this research is to incorporate “real world” constraints such as manufacturability and robustness into the design of spatial linkages. Current research in this field does not fully address these issues. The additional constraints will need to be formulated mathematically so that the space of possible mechanisms can be searched for the optimal solution.  For example, when selecting a motor it is important to balance the benefits of high torque, such as performance time, with the drawbacks such as higher cost and weight.  The exact formulation of this problem is what needs to be determined.

The practical benefit is to provide an alternative to a full robot in complex automated tasks.  The low degree-of-freedom mechanism can be built cheaper, perform as well and have the extra benefit of requiring less power to run it.

A PR device is capable of producing spatial (robot-like) manipulation through the use of far fewer motors than a robot. Being both mechanically simpler and requiring less actuation than a robot, these mechanisms have the considerable potential to produce significantly less waste and operate at higher speeds. With the growing focus on lean manufacturing and its concern for waste minimization, this research focuses on the design and implementation of manipulation systems based on high speed PR devices that are in line with these lean manufacturing goals. This research is of both theoretical and practical significance. The theoretical aspect of the work is the derivation of a methodology that results in viable mechanical (kinematic) layouts for the machines proposed. The practical aspect is the ability to design and implement into assembly lines actual devices based on the derived mechanical layouts. An additional goal of the research is the ability to design and implement reconfigurable devices to allow for some degree of flexibility in their manipulation capabilities. This property should allow these devices to act as part of modern assembly lines capable of handling classes of parts rather than single parts.

Kinematic Synthesis of Spatial Mechanisms and Platforms

Robots are often considered the only mechanical system available for creating complex three-dimensional trajectories or obtaining multiple orientations of a part. A spatial mechanism, however, is an articulated device that can produce movement as sophisticated as a robot. One major difference between a robot and a spatial mechanism is that the latter is a single task device, tailored to meet the specific requirements of the system. These mechanisms can provide precise, repeatable and high strength movement of a part, often accomplishing this through the use of just a single actuator.

Few spatial mechanisms are in use. The conception, design and development of these mechanisms is extremely challenging due to the three-dimensionality of the design process. Currently, few design tools are available for considering spatial mechanisms as an option to more traditional solutions. Research is being conducted in this area that includes the development of design theory, the implementation of this theory in three-dimensional designer oriented graphics environments, and the prototyping of these new devices. The goal of this research is to produce an efficient design and construction methodology for several classes of planar and spatial mechanisms.

This material is based upon work supported by the National Science Foundation under Grant No. 0422731.

This site was last updated 09/16/05