当机器人能够自学……

A 5-year-old can tie their shoelaces, but robot hands aren’t nearly so nimble.

五岁大的小孩已经可以自己系好鞋带子了,而机器人手却远没有如此灵活。

A new system, however, has greatly improved their dexterity.

不过,一种新的系统已经大大提升了它们的敏捷度。

Hard-coding a robot to coordinate multiple joints is daunting.

硬编码一个机器人让协调多个关节,这个任务是艰巨的。

So computer scientists have turned to machine learning, a field of artificial intelligence (AI) in which computers build skills on their own.

于是计算机科学家们转向了机器学习,这个领域指的是人工智能(AI)能够自己学习掌握技巧。

Such learning takes time and repetition, however, and robot hardware is slow and breakable.

不过,这样的学习需要花费时间和反复练习,而机器人的硬件又是迟缓而易碎的。

Some researchers instead train algorithms with virtual robots, but reality is always slightly different from simulation.

于是有些研究者们转而用虚拟机器人来训练算法,不过现实总是跟仿真有轻微的出入。

The new work overcame this “reality gap” by slightly randomizing elements of the simulation during training, such as friction and object size.

这次的新研究则跨越了这道“现实鸿沟”,他们通过将训练中的一些仿真因素轻微随机化来实现,例如摩擦力和物体大小。

(Most of the work, in both simulation and reality, was done with a child’s building block with letters on its sides.)

(在模拟和现实中,大部分工作都是在孩子的积木上完成的,积木两边都有字母。)

They also gave the program short-term memory, so after a few seconds of handling the cube, it got a sense of the block’s exact size and other factors and adjusted for them.

他们还给这个项目的机器人带来了短期记忆功能,所以在控制好方块几秒钟猴,它就会感知到方块的具体大小和其他因素,然后调整适应它们。

The researchers used the commercial Shadow Dexterous Hand, which resembles a human hand, attached to a wall, along with a digital simulation of the hand for training.

研究员们用的是这种商业化的影随学习灵活手,长得就像人手一样,连接到一面墙上,并带有用来训练用的数字仿真系统。

In both virtual training and a physical test to see how well the training transferred to the real hand, the hand was instructed to manipulate a cube in a series of new orientations so that, for example, the side with the A on it was facing up and side with the P on it was facing out.

在虚拟训练和用来看训练传递到真实的机器手的实际测试中,这只手被指示操纵一个方块来面向不同的朝向,比如,有A的一面朝上,而有P的一面朝外。

No robot hand had ever done something nearly as complicated.

之前还没有机器人手能够完成这样复杂的动作。

In the real world, the system “saw” the cube using three cameras placed above the hand.

在现实世界中,这个系统可以利用置于手上方的三部相机“看到”方块。

The virtual hand, after the equivalent of 100 years of trial-and-error practice (sped up in simulation), performed an average of 30 consecutive reorientations without getting stuck or dropping the cube.

而虚拟手在相当于一百年的试错练习(在仿真中加速了)之后,能够进行平均30次连续的重新取向而不会产生卡顿或者把方块弄掉。

The physical hand completed an average of 15 consecutive reorientations without getting stuck or dropping the cube, the researchers report today.

而实际的手可以完成平均15次连续的重新取向而不产生卡顿或者把方块弄掉,研究员们今天发表了结果。

The system, called Dactyl, also discovered common human tricks such as spinning the cube between two fingertips or taking advantage of gravity to shift the block.

这个系统叫做达克提利,也能够玩一些人类的把戏比如在两个手指尖旋转方块或者利用重力来转换积木。

The advance might improve the assembly of delicate electronics or the ability of health care or domestic robots to help around the house.

这些进步也许可以用于提升精密的电子组装,医疗,或者帮忙做家务的家用机器人。

Omelet, anyone?

有人要吃煎蛋卷吗?

关于这种机器人手朋友们有什么想法吗?欢迎给amber留言哦!

感谢您的关注!

跟amber一起

看世界

世界那么大,一起去看看吧!

  • 发表于:
  • 原文链接https://kuaibao.qq.com/s/20180801G0FT5N00?refer=cp_1026
  • 腾讯「腾讯云开发者社区」是腾讯内容开放平台帐号(企鹅号)传播渠道之一,根据《腾讯内容开放平台服务协议》转载发布内容。
  • 如有侵权,请联系 cloudcommunity@tencent.com 删除。

同媒体快讯

扫码关注腾讯云开发者

领取腾讯云代金券