I-Corps: Robotic Hands for warehousing & Automation

自动产品处理的典型客户是具有各种形状和尺寸的高吞吐量产品的仓储业务。这些仓库为消费电子产品、食品分销和药品等领域的在线业务提供存储和履行物流。示例包括Amazon、Target和CVS。大约有17.2万名仓库工人从事手工搬运库存。这些仓库面临着持续但波动的运输订单需求，可能受到季节性假期、时尚趋势甚至一周中的某一天的影响。目前这些仓库的解决方案是在需求增长时雇用额外的员工来挑选和包装这些包裹。将这些订单挑选任务自动化还可以解决季节性劳动力波动和恶劣工作条件的问题。这个研究小组开发了一种专利机器人手，它将彻底改变拣选和包装。拟议中的技术将通过更快地包装和运输产品来创造价值，同时消除员工在压力环境中因重复性工作而受伤的问题。机械手技术的能力是通过两个关键的进步实现的：自适应机械设计和高可靠性触觉传感器。手的机制是基于特殊设计的具有多个自由度的弯曲关节，使手指能够被动地根据物体形成自己的形状，简化控制，减少掌握物体所需的信息，降低成本，并提供柔和的触摸。触觉感应技术帮助手适应物体形状（特别是轻物体），并检测错误。这种设备比竞争对手的技术更敏感，制造起来也容易得多。该团队将致力于为产品处理市场开发这种机器人手技术。目前处理物体的黄金标准是使用真空夹具。这些夹具在设计和实现上都很简单，但它们可以处理的产品范围相当有限。此外，最先进的自动化系统使用专门的夹具，如平行颚式夹具，只能执行狭窄范围的任务，需要专家选择合适的夹具（一个昂贵和缓慢的过程），如果产品线改变，必须更换。

A typical customer for automatic product handling is a warehousing business with high throughput of products with a wide range of shapes and sizes. These warehouses provide storage and fulfillment logistics for online businesses operating in areas such as consumer electronics, food distribution, and pharmaceuticals. Examples include Amazon, Target, and CVS. There are approximately 172,000 warehouse workers involved in moving inventory by hand. These warehouses face a constant but fluctuating demand for shipping orders that can be influenced by seasonal holidays, fashion trends or even day of the week. The current solution for these warehouses is to hire additional staff to pick-and-pack these packages when the demand grows. Automating these order-picking tasks would also solve issues of seasonal labor fluctuations, and the poor working conditions. This research team has developed a developed a patented robotic hand that will revolutionize picking and packing. The proposed technology will create value by packing and shipping products faster while eliminating employee injuries related to repetitive tasks in stressful environments.The capability of the robot hand technology is enabled via two key advances: adaptive mechanical design and high reliability tactile sensors. The mechanics of the hand is based on specially designed flexure joints with multiple degrees of freedom that enable the fingers to passively shape themselves to the object, simplifying control, reducing the information required to grasp an object, lowering costs, and providing a gentle touch. The tactile sensing technology helps the hand adapt to object shape (especially for light objects), and detect errors. The devices are more sensitive and far easier to manufacture than competing technologies. The team will be working on developing this robotic hand technology for the product handling market. The current gold standard for handling objects is by using vacuum grippers. These grippers are simple in design and implementation, but the range of the products they can handle is quite limited. Additionally, state-of-the-art automation systems use specialized grippers, such as parallel jaw grippers, that only perform a narrow range of tasks, require experts to select the proper gripper (a costly and slow process), and must be replaced if product lines change.