What Could Go Wrong? Crime Fighting Robots Now Equipped with Self-Defense Instincts

Free Thought Project – by Claire Bernish

Once thought to be too approachable for their own good — as in, able to efficaciously perform the tasks they’ve been designed to do — robots are now being programmed with self-defense capabilities.

Now, robots which, say, patrol for criminal or suspicious activity come with a blend of humanoid characteristics and self-defense programs which prevent them from being perceived as too cute — or too menacing.  

“Because of all the doomsday scenarios people imagine with robots, their makers have to insert some cuteness,” explained Golden Krishna, a designer with Alphabet’s Google, as reported by the Wall Street Journal.

“When humans see a robot that doesn’t have eyes,” said Rodolphe Gelin, SoftBank Group Corp. robotics unit chief scientific officer, “they think it doesn’t care about them.”

That perception — however Orwellian it might sound — can be detrimental to a robot’s ability to function on the job in the presence of humans.

After several failed design attempts to find the appropriate balance between relatable and threatening — and a number of incidents in which robots faced ‘abuse’ or ‘assault’ by humans — newer robots employ self-defense capabilities.

One incident of, well, ‘roboticide’ occurred with the failed experiment of HitchBOT. This cylindrical robot with an LED smile was dropped off on the roadside to see how if robots could “depend on people”  — but the experiment came to an abrupt halt when HitchBOT “was found decapitated in Philadelphia last summer,” the WSJ noted.

“While the project showed that they can indeed trust humans,” HitchBOT’s creators said generally of robots, “there are also exceptions to the rule.”

Experts on the interaction between humans and their new mechanical counterparts observed “robot abuse” in action in Okinawa, Japan, in 2014, when robots designed to assist the elderly in purchasing groceries came under attack by children, who kicked it, beat it with a bottle and bent its neck.

“The robot said ‘Someone help me!’ and Ouch, that hurt,” said researcher, Dražen Bršcic, with ATR Intelligent Robotics and Communications Laboratories, as cited by the WSJ. “But it didn’t stop the children.”

Those experiments led, in part, to robots with human-like facial features that can also employ defensive tactics if someone gets spooked or wishes to cause harm.

Instead of reacting as if they’re being hurt, some newer robots have been designed to simply stop moving altogether — in theory, to cause attackers to lose interest. Thus, the K5 — a robot who performs all the functions of security patrols except offense — was born.

The WSJ described the K5:

“The camera ended up at nose position. A row of ultrasonic sensors, to help it avoid running into people, looks like buttons. A navigation laser resembles a hat. Curved lights are eye-like.

“The K5 patrols with buddy bots. If someone messes with one, its partner can document the offense. The K5’s shape and 300 pounds offer protection: The lack of edges make it hard for bad guys to lift and steal. Knightscope rents a K5 for $7 and hour or about $60,000 a year.”

This 5-foot, 3-inch, glossy-white, graffiti-resitant K5 now appears almost neutral and can interact with humans more readily than some previous designs — such as a robot resembling a small tank — but is built with the ‘play dead’ reflex if attacked.

Knightscope, the start-up responsible for the K5, sought to create a beat cop robot; and the company’s vice president of marketing and sales, Stacy Stephens — as a former Dallas police officer — has offered input for the project.

“When it comes to mischievous activities,” she said in the WSJ, “I know a bit about what people do.

While the world might not be quite ready for armed robots-on-patrol, it would be a question worth considering as designs improve and people become familiarized to their presence. Armed, unmanned aerial vehicles — drones — have been responsible for a highly-contentious number of deaths in conflicts around the world, and particularly, the Middle East.

Can armed, unmanned, non-aerial, ‘human-esque’ robots really be that far in our future?

Below is a video of the K5 in action.

Read more at http://thefreethoughtproject.com/wrong-crime-fighting-robots-equipped-self-defense-instincts/#rzsGu8FBpjmdbsvd.99

4 thoughts on “What Could Go Wrong? Crime Fighting Robots Now Equipped with Self-Defense Instincts

  1. We’re being introduced to robots slowly via the media, always as friendly and helpful devices, of course, but I’m willing to bet that there’s a billionaire somewhere who has already amassed an army of Robo-cops that he’s just itching to unleash on those pesky poor people.

    SEVERAL YEARS AGO the Japanese showed us an entire orchestra of robots. They have a robot conductor, robot musicians, and if a robot can play the violin, he has all the manual dexterity needed to aim and shoot a rifle. It’s only a change in his programming that turns him from a violin player to a soldier.

    There will be a world war, civil wars, race wars, border wars, and revolutions, because the first goal is to kill as many humans as possible, but anyone who survives all of that, better know how to kill these things too.

    1. Robots will pose a threat, but they’ll have their weaknesses. The more complex any device is, the more there is to go wrong, and the more skilled maintenance it will need. Also, complex robots (like Atlas) aren’t well-suited for cheap, reliable mass production the way some things are (like computer chips).

      Nevertheless, it’s dangerous to depend on optimism. Homemade AP rifle ammo really needs to become a cottage industry, as there just isn’t enough true AP ammo in circulation. I also wouldn’t mind seeing more robotics gurus on our side of the liberty-tyranny fence.

Join the Conversation

Your email address will not be published. Required fields are marked *


*