This Small Business Innovation Research (SBIR) Phase I project proposes a robotic grasper based on a novel torque switching mechanism, and a structurally compliant finger with embedded sensing. Robotic hands in industry tend to be fragile and lack the dexterity to perform a wide range of grasping and manipulation tasks. Robotic hands in academia tend to be more dexterous yet tend to be bulky, possessing large controller cabinets and/or forearms. Furthermore, when a collision occurs within a robotic workcell it is usually the hand (gripper) that makes impact. The proposed hand will have greater dexterity than a conventional gripper, being able to perform both power grasps and pinch grasps on large and small objects and yet be able to absorb and survive major collisions fully intact while alerting the robot system to move more cautiously. The Phase-I project objectives are: (1) create prototypes of an active torqueswitching mechanism; (2) create prototypes of a resilient polymer link with an embedded sensor; (3) build a prototype 2-fingered hand; (4) evaluate functionality of the prototype hand. Phase II will result in a compact, durable hand capable of grasping and manipulating a large range of objects.
The broader impact/commercial potential of this project is three-fold; there will be a general commercial impact, a direct commercial impact, and a societal impact. The general commercial impact addresses the shortcomings of metal gripper devices available today. These gripper devices discourage the use of robotic systems, thereby undermining manufacturing productivity. The proposed solution improves manufacturing competitiveness by enabling easier adoption of robotic work cells. The direct commercial impact is that local (US) workers will assemble, test, and ship these products. Ninety-five percent of the components will be sourced from US fabricators and OEM suppliers, boosting the US economy. Meanwhile, the company exports 50% of its finished products to foreign markets. The societal impact stems from the idea that the risk of damage to an expensive robot discourages innovation and keeps the programming to a select few programming ?experts?. The proposed solution invites production line workers back into the process. Without the high cost of collisions, the worker (true expert) is encouraged to experiment and program process improvements. Corporations otherwise automate these workers out of their skilled jobs who then join the unemployed while the company literally loses touch with the ability to understand and innovate processes.
This Small Business Innovation Research Phase I enabled the introduction of a TorqueMultiplexer™ that actively channels motor torque to different robotic-hand joints from a single motor. The TorqueMultiplexer measures less than 1 cubic centimeter in volume and is capable of transmitting up to 200 mN-m of torque. Furthermore, it requires only 2 Watts of power for the fraction of a second that it needs to engage and subsequently draws no power while remaining engaged. In its present application, this novel mechanism enables dependant and independent (time-discreet) control of two or more degrees-of-freedom using a single motor. This system was installed in a BarrettHand™ and enabled independent control of both the proximal and distal finger joints to allow the hand to perform different types of grasps, such as a power grasp for picking up large objects and a pinch grasp for picking up small objects (between the fingertips). Other innovative elements included the manufacture of a resilient (elastomer) finger link with an embedded sensor that reduces the complexity of the assembly while protecting the electronics. The inherent flexibility of the link means it is less prone to damage from external forces. Finally, we used Barrett’s Puck™ module (a miniature electronics package that controls brushless DC motors) to: (1) control the drive motor, (2) control the TorqueMultiplexer, and (3) to operate the embedded sensor. While several technical challenges remain before the TorqueMultiplexer can be applied in mission-critical commercial applications, the Phase-I results hold promise for a less expensive yet more dexterous robotic hand than the current market leader (BarrettHand).
