Robots have a defined role in our future, namely when it comes to aiding humans in tasks ranging from routine errands and chores, to running complex industrial facilities and machinery. While integrating some degree of independence via artificial intelligence (AI) is key in the efforts of getting robots to reach this degree of operability, this won’t be possible without assuring these droids are safe to use.
A new report that emerged from Seattle-based cybersecurity firm IOActive has shed some light on how potentially difficult this could wind up being. The report demonstrated multiple ways cybercriminals can infiltrate a number of popular robots used for practical and complex tasks. The report goes into detail on ways these machines can be turned into surveillance devices that send audio and video of their owners back to hackers, along with how these robots can be remotely controlled in ways that could be harmful to humans.
As the video below shows, hackers are capable of making humanoid robots perform gestures and motions that could be perceived as dangerous or aggressive.
Researchers were able to breach notable robotics brands like Pepper, the Alpha Robots, and Nao. Granted a robot like the one shown in the video won’t cause overly devastating damage, but that might not be the case if (or when) larger, more intelligent humanoid robots become publically available, or if hackers gain control over industrial robotic equipment.
It’s worth noting home robots like the Alpha 2 are becoming more powerful and capable of human-like movements and gestures as technology continues to improve. IOActive even proved that those bigger industrial robots mentioned earlier aren’t impervious to cyberattacks. Having said that, IOActive researchers showed they were capable of compromising industrial robot arms made by a company called Universal Robotics. These limbs were designed to work alongside humans, but researchers found ways to override their safety protocols. This required access (or at least be able to physically tamper) to the same network as these robots.
Obviously, being able to control such a bot could have disastrous consequences. Universal Robotics’ creations have enough power to deliver a force that’s more than sufficient to cause a skull fracture—even while running at low speeds. These kinds of reports don’t necessarily come as a surprise, but prove how we continue to take the security of our connected devices for granted. Last year, an array of infiltrated IoT devices like cameras, lightbulbs, thermostats, and others were formed into a giant botnet that attempted to take out the Internet.
Imagine the damage an equivalent assembly of actual humanoid robots armed with weapons and enhanced physical capabilities could do under those same circumstances?
Filed Under: Cybersecurity, M2M (machine to machine)