权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Provocation of robot behaviors based on the affordances of mechanical objects

基于机械物体的可供性激发机器人行为

基本信息

批准号：
11650265
负责人：
TANI Kazuo
金额：
$ 2.24万
依托单位：
Gifu University
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
1999
资助国家：
日本
起止时间：
1999 至 2000
项目状态：
已结题

来源：
https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-11650265/
关键词：
affordance robot autonomous machine behavior disassembly of product

项目摘要

Simulations have been conducted to show the possibility of mechanical products to be disassembled by a robot that finds parts in the product and provokes behaviors proper to the parts. A behavior of turning over the product is added to the conventional behaviors of turning screws and taking parts apart to enables the robot to attack multiple sides of the product. The robot's provocation of the behaviors can be regarded as the robot's acceptance of the affordances presented by the product or parts. We have shown by simulation that imposing certain rules to the acceptance of affordances, i.e. provocation of behaviors, makes the disassembly proceed effectively. For example, the robot should try another object if it fails in taking one apart, and it should try on another face of the product if it fails on one.To demonstrate this concept, we have tried disassembly of real but simple mechanical products and built a network robotics system consisting of several subsystems, each taking care of … More its own device such as a robot, a vision sensor, a force sensor, etc., to carry out disassembly experiments. The vision subsystem detects hexagonal and rectangular shapes, corresponding to bolts and blocks, and sends their positions to the robot subsystem. The robot subsystem, for example, provokes the behavior of turning counterclockwise in case of a hexagon, and the behavior of pulling upwards in case of a rectangle. We use such methods as template matching, autocorrelation, and Hough transformation.To investigate the behavior to remove a part using force sensor information, a structure of behaviors for escaping a maze using force sensor information is built referencing to a human behavior.For the cooperation between robots, an experiment is conducted on the network robotics system, where the motion of a robot is measured by a position sensor subsystem for another robot of another subsystem to follow.Thus the fundamentals for the experiment of the provocation of robot behaviors based on the affordances are established. Less

仿真显示了机器人在产品中找到零件并激发零件的适当行为来拆卸机械产品的可能性。在常规的拧螺丝和拆件行为的基础上增加了翻转产品的行为，使机器人能够对产品的多个侧面进行攻击。机器人对这些行为的挑衅可以看作是机器人对产品或部件所提供的启示的接受。我们已经通过模拟表明，对允诺的接受施加一定的规则，即行为的挑衅，使拆卸有效地进行。例如，如果机器人无法拆卸一个物体，它应该尝试另一个物体；如果它无法拆卸一个物体，它应该尝试产品的另一个表面。为了证明这一概念，我们尝试对真实但简单的机械产品进行拆卸，并构建了一个由几个子系统组成的网络机器人系统，每个子系统都有自己的设备，如机器人、视觉传感器、力传感器等，进行拆卸实验。视觉子系统检测六角形和矩形形状，对应螺栓和块，并将其位置发送给机器人子系统。例如，机器人子系统在六边形的情况下激发逆时针旋转的行为，在矩形的情况下激发向上拉的行为。我们使用了模板匹配、自相关和霍夫变换等方法。为了研究利用力传感器信息移除零件的行为，参考人类行为，构建了利用力传感器信息逃离迷宫的行为结构。对于机器人之间的协作，在网络机器人系统上进行了实验，其中一个机器人的运动由一个位置传感器子系统测量，另一个子系统的另一个机器人跟随。从而为基于启示的机器人行为激发实验奠定了基础。少