This Week In Bots: More Human Now Than Machine

Among the many advances in robot science comes a surprise: Our brains may think of even the most machine-like robots as slightly human.

This Week In Bots: More Human Now Than Machine

Bot Vid: Hopping For A Better Robot

Hopping robots make use of the relative efficiency of hopping versus other modes of locomotion, especially flying, but as the Automaton blog points out Georgia Tech researchers have just come across a clever way to make hopping bots ten times more efficient still. It’s simply, it seems: A tiny jump timed precisely just before a bigger leap. The stutter disrupts the energy store/release cycle of the hopping action in a way that means much less energy is required to reach the same hop height. Simple (but painstakingly researched) innovations like these are the sort that can transform near-future robotics.


Bot Vid: Photobot

Robots are mechanical contrivances designed to replicate the work of a person, and in this case PhotoBot is a little machine that replaces a photographer at your party. Coming from a postgrad at Northumbria University it’s supposed to be set up at an event where it then scans around automatically snapping photos of people in action. Cleverly it uses ultrasonic ranging to ensure it can snap away accurately, and it has a slightly anthropomorphized “face” to make it seem cuter. Though it’s not likely to turn into a buyable product, it’s a taster of future robots that will help you do your own life-casting photography. And, inevitably, of surveillance robots.

Bot Vid: Korean Chef Bot

The battle to build a better butler bot bumbles on, seemingly ahead of many other branches of robot development (are we really that desperately hunting for home-helper robots?), and the latest moves come from the Korean Institute of Science and Tech. Their latest innovation with their CIROS robot is to give it enough skills to cut and prepare food for a salad and even pour drinks and do dishes. It’s clever enough to recognize kitchen equipment and it’s another hint that your retirement relaxation may be assisted by a home robot.

Bot News

Waveglider survives Sandy. A plucky little seagoing robot from Liquid Robotics, named Mercury, hit the headlines this week because it survived the brunt of hurricane Sandy and actually measured the storm’s weather parameters as it hit. It’s not surprising, given the relatively small superstructure the robot has and its strong, shallow sub-surface chassis, but it’s a testament to the power and usefulness of robots nowadays.

Bionic leg climbing in Chicago. This weekend amputee Zac Vawter is scheduled to do something rather impressive: He’s going to climb all 103 flights of the Willis Tower building in Chicago. The feat is possible thanks to his thought-controlled bionic leg, and as well as a testament to Zac’s will power and a stunt to raise money for and awareness of the Rehabilitation Institute of Chicago, it’s a reminder that robotics research can have very human pay-offs.

Medical robotics advances. We hear a lot about use of surgical robots nowadays, thanks to their precision and faster patient recovery post-op times, but news from Italy is even more intriguing. The team here is working on a prototype robot that is partly assembled inside the patient, after being inserted through a small incision in the navel. From here its controlled remotely by the surgeon, and future robots like this may have more dexterity and thus surgical potential than existing external small-port surgical bots like da Vinci.


Robot Vs. human car drivers. The BBC reports on a race between a human-driven car and a robot-piloted car that just happened at the Thunderhill Raceway in California. While the human pilot, already familiar with the track, actually won the race it was by the narrowest of margins. The car, made by Stanford’s Center for Automotive Research, is a hint that self-piloting cars may be on the roads sooner rather than later, and their prowess in driving may allow for efficiency and safety may soon exceed what human drivers can do.

Bot Futures: Here’s Looking at You

I suppose when we think of real-life robots, we have a tendency to consider them implacable pieces of hardware. Sure, Asimo’s antics are impressive, and his child-like size and expression are slightly endearing…but you’d hardly look at a telepresence robot with the same sort of emotion. That’s in contrast to the robots we imagine in movies and TV shows, where–through the usual magic of suspension-of-disbelief–we can feel genuine attachment to fictional robots, even non-anthropic ones like the adorable trio in the movie Silent Running. But research from New Zealand is suggesting something rather different, and it may be a testament to our already increasingly robotized society.

University of Canterbury researchers have been studying how people’s brains react when they perceive an object like a robot. The study was based on the psychological phenomenon of the inversion effect–the difficulty we have recognizing some objects when we see an image of them upside down compared to other objects. Faces or human body postures, for example, are harder to recognize this way, and it’s commonly thought that it’s because our brains are pre-coded to recognize other people and to try to interpret their bodies than other non-human objects.

The researchers simply tried to see if their human subjects could recognize images of upside-down and right-way-up robots in the same way as they saw other humans or in the way they identified inanimate objects. While we may expect, like the research team did at first, that people may at least identify more human-looking robots (think of C-3PO right now) in the same way we spot human bodies, it turns out that all sorts of different robots, even machine-looking ones, were perceived as more human-like than we perceive other objects (i.e., they were perceived with a stronger inversion effect).

That’s odd, perhaps. But the research team interpreted it as a sign that when we look at robots, some concept associated with them activates very deeply ingrained brain functions that we otherwise use to spot people. As well as telling us that something fascinating is going on in our brains, the team suggests that the conclusion may influence future robot design because on the whole we may treat robots more like people.


That at least explains a report this week in the Pacific Standard that people are being observed as becoming emotionally attached to their social care robots, now increasingly used in medical situations in some parts of the world. Scientists have found cases of people kissing their robot companions, and even confiding in them or buying them gifts as they would a friend. A broken robot can even, it seems, lead to a broken heart. And, if you think about how attached you are to your smartphone (be honest–it’s for a little more reason than its expense or the fact it contains so much personal data) it’s reasonable to expect that a robot you spend a lot of time with may become even more emotionally valuable.

Just think about that for a moment, and then remember how nice it was to read, up above, that a plucky little robot managed to make it alone through the stormy seas beneath hurricane Sandy.

[Image: Flickr user jurvetson]

Chat about this news with Kit Eaton on Twitter and Fast Company too.

About the author

I'm covering the science/tech/generally-exciting-and-innovative beat for Fast Company. Follow me on Twitter, or Google+ and you'll hear tons of interesting stuff, I promise. I've also got a PhD, and worked in such roles as professional scientist and theater technician...thankfully avoiding jobs like bodyguard and chicken shed-cleaner (bonus points if you get that reference!)