December 12, 2014 06:18am ETRobot Swarm! NYC Exhibit Uses Bots to Teach Math  MoMath's newest exhibit, "Robot Swarm," features two dozen of these small, wheeled robots. Credit: Meuseum of Math View full size image

NEW YORK — A new interactive exhibit in New York City teaches kids and adults alike about the mathematical order of the natural world in an unconventional way: with dozens of swarming robots.

At first glance, the "Robot Swarm" exhibit — which opens Sunday (Dec. 14) here at the Museum of Mathematics (MoMath) in New York City — looks more like a futuristic boxing ring than a museum display. Essentially, it's an elevated box cordoned off by thick, metallic rope. At a preview of the exhibit, three people in backpacks (who happen to be MoMath's co-founders and chief designer) meander around the ring, performing what appear to be fancy footwork and exchanging mild smack talk. But no one throws any punches.

As the members of the trio move around, each is followed by a pack of tiny robots that roll around right under their feet. The robots, which look like an army of mechanical horseshoe crabs, are clearly visible through the exhibit's transparent floor. [The 6 Strangest Robots Ever Created]

MoMath co-founders Glen Whitney and Cindy Lawrence play around inside the museum's new "Robot Swarm" exhibit. Credit: LiveScienceView full size image

The bots flash red, green or yellow depending on which backpack-clad human they're following at the moment. The mechanical army is in "pursue-mode." Each backpack contains a sensor that allows the robots to detect the location of the wearer. Once detected, the wearers are swarmed.

"In a swarm, the whole is more than the sum of its parts. There's almost no individual intelligence, but [the individuals] create this group intelligence because of the interaction of their behaviors," Glen Whitney, MoMath's other co-founder and co-director, told reporters Wednesday (Dec. 10) at a preview of the exhibit.

To make sure the individual robots' behaviors are part of a bigger picture, each bot maintains radio communication with a central computer, which dictates which of five different behaviors the bots should perform. At this week's demonstration, the bots pursue the people in the ring, but when in "run away" mode, the tables get turned, and the robots flee from the people in the ring. In "robophobia" mode, the robots flee each other, with each bot trying to get as far away as possible from its comrades.

All of these behaviors are examples of what Cindy Lawrence, MoMath's co-founder and co-executive director, called emergent behavior — a mathematical concept that helps explain how simple, local interactions can lead to large-scale organized behavior. You can see this concept at work in the exhibit, where robots appear to be carrying out some complex plan but are really just following one global rule, Lawrence said.

For example, when the bots swarm around Lawrence's feet, the rule they're following is simple: get as close to the sensor as possible.

In the real world, many tasks performed by robots are (or soon will be) aided by an understanding of emergent behavior. The ultimate goal for those who study robots in this context is to understand the relationship between the bots' simple, local interactions and their complex group behaviors, said James McLurkin, a professor of computer science at Rice University in Texas, and one of MoMath's roboticists in residence.

"The Holy Grail is to identify some global goal and then somehow get all these robots to get it done," McLurkin told Live Science. "And you, the human, never have to specify the actions of each individual robot."

McLurkin helped MoMath get 24 bots to accomplish global goals for the Robot Swarm exhibit, and he's done the same thing with at least 100 robots in his lab at Rice. Michael Rubenstein, a postdoctoral fellow at Harvard University, has figured out how to get 1,000 little robots to perform simple group tasks, such as moving around to form a particular shape.

In the natural world, thousands of creatures coordinate more complex behaviors than Rubenstein's bots perform — like avoiding a predator, or building a hive. Fish, honeybees, wolves and geese are some of the many animals that exhibit individual behaviors that allow them to act in concert with their peers, Lawrence said. And it was these examples from the natural world that inspired Robot Swarm.

"Math problems in schools don't always feel realistic to kids," Lawrence told Live Science. "We want kids to see that math has a relationship to the natural world. Math is all around us."

Robot Swarm opens to the public on Sunday (Dec. 14).

机器人群！纽约展览馆用机器人教数学

Robot群！纽约展览使用机器人来教Math  MoMath最新的展览，“机器人群，”拥有两打这些小，轮式机器人。

NEW YORK＆mdash;在纽约市的一个新的交互式展览教孩子和成人一样对自然界的一种非常规的方式的数学顺序：与几十个蜂拥机器人.

乍一看，在与;机器人群;展览;这将打开周日（12月14日）在这里数学博物馆（MoMath）在纽约市和mdash;看起来更像是一个未来的拳击环比博物馆展出。从本质上讲，它是一个高的中厚，金属绳子拉起了警戒线。在展品预览，三个人在背包（谁碰巧MoMath的共同创始人和首席设计师）蜿蜒周围环，执行什么似乎是看中步法和交换轻度嫌谈话。但是，没有人抛出软柿子.

作为三重奏的成员走动，每个后面那打滚权在他们脚下的小机器人一包。机器人，看起来像机械马蹄蟹的军队，显然是通过展览的透明地板可见。 [6奇怪的机器人有史以来The 6 Strangest Robots Ever Created

的机器人闪烁红色，绿色或黄色取决于他们在之后的时刻在其背包包的人。机械部队是在与;追求模式和。每个背包包含一个传感器，其允许机器人以检测穿用者的位置。一旦检测到，配戴者都蜂拥.

;在一个群，整体大于各部分之和。有几乎没有个人的智力，但个体the individuals格伦·惠特尼，MoMath的另一位创始人和联席主任告诉记者，周三（12月10日）在展览.

为了确保各个机器人的行为是一个更大的画面的一部分预览，每个机器人保持无线电通信与中央计算机，这决定了它的五个不同行为的机器人应该执行。在本周的演示中，机器人追求的人在擂台上，但是当与私奔;模式，表得到转过身来，机器人与人在擂台逃跑。在模式，机器人逃离对方，每个机器人试图得到尽可能远地从它的同志.

所有这些行为都是什么辛迪·劳伦斯，MoMath的联合创始人和联合执行董事，被称为自发行为和mdash例子;一个数学概念，有助于解释如何简单的，局部的相互作用可能导致大规模的有组织的行为。你可以看到在展览，其中机器人似乎在进行一些复杂的计划，但实际上只是下面的一个全局规则这个概念在工作中，劳伦斯说.

例如，当机器人周围劳伦斯的双脚蜂拥而上，规则，他们“再下面是简单的：获得尽可能接近传感器尽可能.

在现实世界中，由机器人进行多任务（或即将）由计算机辅助的自发行为的理解。对于那些谁研究机器人在这种情况下的最终目的是要了解机器人“简单，当地的相互作用及其复杂的群体行为之间的关系，詹姆斯说McLurkin，计算机科学教授莱斯大学的得克萨斯州，以及MoMath的机器人专家1在居住.

u0026 QUOT;圣杯是确定一些全球性的目标，然后以某种方式得到所有这些机器人来完成它， McLurkin告诉现场科学。 ＆QUOT;而你，人，永远不必指定每个机器人的动作在他的实验室在水稻。迈克尔·鲁宾斯坦，博士后在哈佛大学，想出了如何让1000的小机器人来执行简单的组任务，如走动，形成特定的形状.

在自然界里，成千上万的生灵协调更复杂的行为比鲁宾斯坦的机器人执行;像避开捕食者，或建立一个蜂巢。鱼类，蜜蜂，狼，鹅是一些许多动物表现出的个人行为，使他们能够在演唱会上与同行行事，劳伦斯说。它是从自然界的启发机器人群.

这些例子，在学校的数学问题，总觉得不切合实际的孩子，劳伦斯告诉现场科学;我们希望孩子们看到数学有与自然世界的关系。数学是我们周围和