Robotic sentience has long been a trope of apocalyptic doomsday enthusiasts – when the machines achieve artificial intelligence, humanity’s days are numbered. Well, it might be time to start letting your loved ones know how much they mean to you: Researchers at Harvard have developed a collection of 1,000 “kilobots” that can work together to carry out basic commands, even overcoming hitches ad roadblocks.

“Biological collectives involve enormous numbers of cooperating entities — whether you think of cells or insects or animals — that together accomplish a single task that is a magnitude beyond the scale of any individual,” said lead author Michael Rubenstein, a research associate at Harvard SEAS and the Wyss Institute.

The kilobots, in many ways, resemble a colony of ants. Once the researchers impart a set of origin coordinates to four main bots, the 996 others receive simple commands that they follow in a variety of ways – following along the edges of the group, tracking their distance from the origin bots, knowing where they are in relation to the objective, etc. The versatility allows the bots to seemingly take on minds of their own to accomplish their goal.

In the demonstrations, researchers put the bots through a variety of maneuvers – they grouped to form the letter “K,” and then reconfigured to form a star shape, for instance. Though similar technology has been implemented in groups of up to 100, this is by far the biggest application. That scale is necessary – the bots were designed to be cheap yet imperfect, and so where a few bots might fail, the sheer number involved ensure success.

“Increasingly, we’re going to see large numbers of robots working together, whether it’s hundreds of robots cooperating to achieve environmental cleanup or a quick disaster response, or millions of self-driving cars on our highways,” said Radhika Nagpal, the Fred Kavli Professor of Computer Science at the Harvard SEAS. “Understanding how to design ‘good’ systems at that scale will be critical.”

For now, the kilobots are relegated to acting as a testing ground for AI algorithms, though more robust applications are sure to follow as researchers begin to understand the real-world dynamics.