Robust Robotic Mechanisms and Sensors via Shape Deposition Manufacturing


Aaron M. Dollar and Robert D. Howe

Support provided by the Office of Naval Research





Project Overview
One of the greatest successes of biologically-inspired design has been the development of mechanically robust robots. One promising biomimetic facbrication technique is Shape Deposition Manufacturing (SDM), which alternates material deposition and machining to produce robot structures with compliant joints and embedded sensing and actuation elements. We explore the benefits of using Shape Deposition Manufacturing for constructing a simple two-fingered gripper and add to the tools available to robot designers by developing a range of sensing modalities compatible with the process. These include Hall-effect sensors for joint angle sensing, embedded strain gauges for 3 axis force measurements, optical reflectance sensors for tactile sensing, and piezoelectric polymers for contact detection. In addition to a simple construction process, the resulting parts are extremely robust, fully functional after high impact loads and other forces due to unintended contact.

[pic]

[pic]
Courtesy of Mark Cutkosky
Manufacturing Process
Shape Deposition Manufacturing (SDM) is a rapid prototyping process involving a cycle of deposition of part material and shaping that builds up the part in distinct layers, resulting in the concurrent manufacture and assembly of the part. In this way, the part can be manufactured in multiple sections or layers, allowing manipulation of the internal parts of the structure. The process has a number of advantages over other prototyping techniques. The deposition of part material allows components to be embedded into the part during production, eliminating the need for fasteners, and reducing the likelihood of damage to the component by encasing it within the part structure. This is a particularly desirable property for the inclusion of fragile components such as sensors, greatly increasing the robustness of the part. Also, depositing the part in layers permits the use of dissimilar materials, allowing for variation of mechanical properties within the same part. This property can be utilized to create complex mechanisms from a single part. Due to its relative simplicity, custom tooling is not required to realize the SDM process. Complex part geometries can be attained using common computer numerical controlled (CNC) mill machines. Look here for more information on the SDM process.


A Robust Compliant Robot Grasper
Robustness is a limiting factor in experimental development of multifingered robot hands: the expense and fragility of these hands precludes casual experimentation, restricting the type of experimental tasks that can be reasonably attempted and slowing implementation due to the need for careful validation of programs. Using polymer-based Shape Deposition Manufacturing, we have produced a novel compliant robotic grasper with the functionality of conventional metal prototypes but with robustness properties that allow for uncertainty in object location and large impact forces. Look here for details.

The compliance and geometric configuation of this grasper is based on the results of previous optimization studies. These studies showed that for a particular set of joint stiffnesses and rest angles, objects could be securely grasped for the widest range of uncertainty in object size and location. Contact forces were also minimized at approximately the same gripper configuration.

[pic]


SDM Embedded Sensors
Shape Deposition Manufacturing (SDM) has shown great promise in enabling biomimetic construction of robust robots that part from traditional design methods. The capability to create spatially-varying materials for complex flexures and the added durability from embedded components will surely prove appealing to researchers frustrated with the fragility and complexity of robots built with traditional manufacturing methods. To further the usefulness of SDM, we have developed sensors for use with the process that cover four of the most utilized sensor types in robotics – an angular sensor for flexural joints, a force sensor with embedded strain gages (in cooperation with Chris Wagner), a compliant tactile sensor, and a low-threshold contact sensor. Results show that SDM can bring the benefits of robustness and simplified construction to a wide variety of robotic sensing applications. Look here for details.



Primary Publications


For more information contact Aaron Dollar, adollar@deas.harvard.edu




Harvard BioRobotics Laboratory Home