未登录,请登录后再发表信息
最新评论 (0)
播放视频

即将到来的机器人族群

The Coming Robot Swarms

成百上千只蚂蚁 即使有两只蚂蚁不小心撞到一起 也不会因此而停下脚步 它们永不止步
Ants don’t get trafficjams. Hundreds of thousands of ants bjjn and the flow doesn’t even stop when two ants bump into each other. They all just keep on moving.
Hhhhh Ghbbh
他们认为我们能从这些小小旅行者 以及其他有着相似行为的动物身上学到点什么
we can learn a thing or two from these tiny travelers and lots of other animals that do similar things.
白蚁殖民地 斑马群 鸟群 鱼群 它们都表现出了某种群体行为 它们可以一起做某件事情
Colonies of termites, herds of zebras, flocks of birds, schools of fish they all exhibit some kind of swarming behavior, And they can do great things together,
比如步调一致地前进
like perfectly synchronizing their movements,
或者建造巨大的巢穴 如果我们能够让机器也这样运作 那不是很棒吗?
or building huge mounds. Wouldn’t it be great if we could get machines to act like that?
使其未来能够在建筑物内的搜救任务中 发挥作用
that in the future might be able to help with everything from building construction to search-and-rescue missions.
群体行为适合应用于 机器人领域的原因是
The thing that makes swarming behavior so perfect for robotics is that in a swarm,
群体中的每个成员都不会进行过于复杂的行为
no one member does anything that’s too complicated.
每个动物只需要遵循简单的规则
An animal is just following a few simple rules,
比如与它周围的每个同伴都保持相等的距离
like staying the same distance from all of its neighbors.
也就是说 你不需要把机器人制造的极其先进
That means you don’t need to make the robot super fancy,
也不需要为每个机器人单独设置指令
and you don’t need to program each one
为其下达精准的指令
to tell it exactly what to do.
相对的你只需要为一批机器人设置同样的基本规则
Instead, you just give a bunch of robots the same basic rules,
因为 无论这些规则是应用于一个多大的群体
and because of how those rules play out in large numbers,
它们都会自行组织 并且判断该做些什么
the group will self-organize, and figure out
不管你要它们去做的事有多么复杂
how to do whatever complicated thing you want them to do.
今天的机器人 已经将其变成了现实
This is already a reality in today’s robotics.
在2014年 哈佛大学的研究人员
In 2014, we told you about some researchers at Harvard
制造出了一千多个这样的机器人
who made over a thousand robots that could arrange themselves
它们能组成科学家们想要的任何形状
into almost any pattern or shape the scientists wanted.
研究人员没有告诉过任何一个机器人该何去何从
The scientists never told any individual robot where to go.
相反 他们只是给每个机器人
Instead, they just gave each one of them
都设置了相同的简单指令
the same simple rules to follow: like,
例如 与周围的机器人保持相同的距离
measure how far you are from your neighbors,
或者 找到机器人群的外边缘
or find an outer edge of your robot swarm,
然后移动到边缘
and move along that edge.
通过反复这些行为
By doing those things over and over,
机器人就可以准确地找到该去的位置
the robots figured out exactly where to go.
哈佛的工程师们也从白蚁身上找到了灵感
Those same Harvard engineers also took some inspiration from termites
他们设计出了一种机器人
to make robots that could build pyramids, castles,
能用泡沫块来建造金字塔 城堡等建筑物
and other structures out of foam blocks.
在这种情况下 他们借鉴了
In this case, they borrowed a strategy
白蚁的一种特殊策略——间接通信
that termites use, known as stigmergy-
这是一种个体间为了达到同一目标
a method of indirectly communicating with each other
而进行间接交流的方法
to reach a common goal.
当我们在做大规模的建设项目时
When humans work on huge construction projects,
我们需要清单 蓝图和指挥系统
we need checklists and blueprints and chains of command.
这些都包括了大量的直接交流
And all that involves a lot of direct communication.
但是白蚁在筑巢时
But termites build by paying attention to tiny clues
会将注意力放在同伴留下的蛛丝马迹上
left over by fellow termites in their environment.
当它们制造小泥块时 会在上面留下自己的费洛蒙
When they make mud balls, they add in some pheromones,
用来提示其他白蚁该去哪里筑巢
which tells other termites where to build.
每只白蚁都在做自己的工作
Each termite is really doing its own thing,
但是这种间接通信
but this indirect form of communication
又使它们之间可以相互协调
allows them to coordinate their actions.
在哈佛大学 研究人员用了一个相似的方法
At Harvard, researchers used a similar idea
使机器人可以根据建筑物当前的形态
to design robots that could place blocks based on
来判断出下一块砖该放到哪里
what the structure looked like at the moment.
也就是说 一个机器人放砖的位置
So one robot could put its block somewhere
可以告诉下一个机器人
that indicated where the robots behind it
该把自己手里的砖放到哪里
should put their own blocks down.
所以 它们并不是在盲目地建造
So, they weren’t just blindly building
因为它们是受程序控制的
because of how they were programmed.
而且 机器人可以进行动态调整
Instead, the robots could adapt on the fly,
就算是研究人员故意干扰它们
even when the researchers tried to mess with them
把它们之前放置的砖移走
by moving blocks that robots had previously put down.
机器人放砖的位置取决于
Each robot placed its block based on
上一块砖在初始位置前是如何摆放的
how the block that was put down before it was oriented.
这些机器人族群到目前还只局限于实验室中
These robot swarms are so far confined to labs,
但是这种想法最终将会服务于我们
but the idea is to eventually have them work for us,
来解决现实世界的问题
and solve real-world problems.
这些间接通信机器人不好的变种
Some variation of these stigmergic robots
可能会用于在危险的地点进行建造
might be able to build things in dangerous places,
比如灾区 抑或是火星
like disaster areas-or even on Mars.
马上 机器人族群就会遍布每个角落
So soon, there might be robot swarms all over the place.
于是我们又要回到堵车的问题了……
…which brings us back to traffic.
科学家已经发现
Scientists have found that
如果把汽车也看做一个族群
car traffic would flow a lot more smoothly
每辆车都遵循着同样简单的规则来行动
if cars acted more like members of swarms,
比如与你周围的车保持同等的距离
with everyone following the same simple rules-
并且让周围的车明白你在干什么
like staying the same distance from your neighbors
那么堵车的问题将会得到很大缓解
and letting nearby cars know what you’re doing.
你可能已经注意到了
Humans, you might have noticed,
我们人类在这一点上确实做的很糟糕
are pretty bad at this.
有一些人甚至都不会使用转向灯
Some people don’t even communicate well enough
给其他司机打信号
to use their turn signal
他们在绿灯时还停在原地
and they sit there at the green light
“因为我根本不知道你在那干什么!在玩愤怒的小鸟吗?”
‘cause, I don’t know, what are you doing? Playing Angry Birds?
但是我们通过基于族群的技术
But we could use swarm-based technology
来开发无人驾驶车辆
to teach self-driving cars
使它们与周围的车保持相同的距离
to stay the same distance from cars around them,
或者还可以让它们
or we could design them
定期地提示路上的其他车辆
so that they periodically let other cars on the road
自己在做什么
know what they’re doing.
像自动汽车一样
Like everything with autonomous vehicles,
这些也都处于测试和开发阶段
it’s all still in the testing and development phase.
但是没准有一天
But the natural logic of swarms
族群的自然逻辑会成为我们解决交通堵塞的关键
might someday be our ticket to less gridlock.
感谢观看这一期的SciShow
Thanks for watching this episode of SciShow,
本节目由我们在网上发起的众筹提供赞助
which was brought to you by our patrons on Patreon.
如果你也想为这个节目出一份力 请访问patreon.com/scishow
If you want to help support this show, just go to patreon.com/scishow.
最后 不要忘了到youtube.com/scishow订阅我们的节目哦!
And don’t forget to go to youtube.com/scishow and subscribe!

发表评论

译制信息
视频概述

动物的族群间交流机制为科学家提供了灵感——我们能否使用这项技术来制造机器人、在特殊环境下修建建筑物、解决交通拥堵?

听录译者

收集自网络

翻译译者

莉萝酱

审核员

自动通过审核

视频来源

https://www.youtube.com/watch?v=uAAYivTtOpQ

相关推荐