The Evolution of Humanoid Robots: Human Reflexes

By Nicholas West

This is an update to the chronicle detailing the evolution of humanoid robots (with videos)

In the age of computers, things evolve exponentially. In just a few generations robots have gone from a scientific fantasy, to a playful curiosity, to entering the battlefield to replace and/or augment their human counterparts.

We are already at the point where we have to consider what the next step of robotic evolution looks like. According to robotics engineers, it appears that at some point in the near future the next step could very well be whatever the next generation robot chooses for itself.

Just late last year it was posited that the humanoid robot was poised to take a leap from a mere facsimile of human behavior to one that futurists suggest will not only walk like a human, but will possess self awareness, as well as a full range of high-tech computational spectrum analysis and capabilities . . . and emotions.

That day has apparently now arrived.

The chronicle below charts the advancement from the rudimentary, through the downright creepy, and toward today where, according to the final video from New Scientist, we see that the new generation of iCub humanoid robot can in fact determine its own goals and exhibit emotional behavior and language skills that so far have been exclusively human.

You will notice that the physicality of robots has now advanced to a point that is leading to serious concern about the economic impact of humans being outsourced to robots for tasks as diverse as service, manufacturing, nursing, housework, yard maintenance and full-fledged agricultural duties. Some are predicting that robots of all types could fully replace humans by 2045.

It is typical of any science with military applications to evolve in a dual- or even multi-use fashion. For example, humanoid robots were initially advertised as a study in how to benefit those who have lost limbs; and, indeed, this has been a noted benefit. However, the real money — the black budget money — goes into applications which can be downright frightening.

A perfect case in point might be MIT’s new HERMES robot, which has an extraordinary level of human flexibility and reaction time. As you’ll see in this video, it is clearly stated that this is an attempt to put a human’s brain inside a robot. But, in the case of HERMES, there also could be a human on the other end operating as a remote-control “pilot” – at least until these robots become fully autonomous.  For the “rescue mission” of course. Or as the soldier, or as the robot police officer, or as….

Enter DARPA, which had already been working on its own projects, but through its Robotics Grand Challenge has spurred a huge influx of inventors looking to receive DARPA funding to offer the next generation of humanoids. The Challenge has been couched in the comforting language of “disaster response” research, but what is emerging seems to be capable of far more than that:

The primary goal of the DARPA Robotics Challenge program is to develop ground robotic capabilities to execute complex tasks in dangerous, degraded, human-engineered environments. The program will focus on robots that can utilize available human tools, ranging from hand tools to vehicles. The program aims to advance the key robotic technologies of supervised autonomy, mounted mobility, dismounted mobility, dexterity, strength, and platform endurance. Supervised autonomy will be developed to allow robot control by non-expert operators, to lower operator workload, and to allow effective operation despite low fidelity (low bandwidth, high latency, intermittent) communications. (Source)

From the task section:

The specific tasks are:

1) The robot will maneuver to a open frame utility vehicle, such as a John Deere Gator or a Polaris Ranger. The robot is to get into the driver’s seat and drive it to a specified location.

2) The robot is to get out of the vehicle, maneuver to a locked door, unlock it with a key, open the door, and go inside.

3) The robot will traverse a 100 meter, rubble strewn hallway.

4) At the end of the hallway, the robot will climb a ladder.

5) The robot will locate a pipe that is leaking a yellow-colored gas (non-toxic, non-corrosive). The robot will then identify a valve that will seal the pipe and actuate that valve, sealing the pipe.

6) The robot will locate a broken pump and replace it.

The robot will be teleoperated, at least at the supervisory level. DARPA will control the communications bandwidth and latency, in order to make the task more difficult and force higher levels of autonomous behavior. If necessary, this control over communications will be used to discriminate performance levels between competitors and select a winner.

[…]

The goal of this Grand Challenge is to create a humanoid robot that can operate in an environment built for people and use tools made for people. The specific challenge is built around an industrial disaster response.

The current array of robots designed over the last few years to match human capability demonstrate what is listed above, and then some:

PETMAN: An anthropomorphic robot designed for testing chemical protection clothing. Its range of motion allows it to balance, walk, crawl, and even climb stairs, while also having the ability to simulate human physiology such as sweating. (Source)

B.E.A.R: Battlefield Extraction Assist Robot. Demonstrates a wide range of capability, while adding a semi-human appearance.

ARM: A new level of robotic hand that employs an opposable thumb to increase dexterity vs. the robot seen above which uses plier-like grips. The ARM is capable of unlocking and opening doors, as well as picking up thin objects like cards or keys. (Source)

Golf robot: This might seem like a goofy novelty, but golf is an extremely precise sport with a lot of room for error — humans can play for decades and never reach a level of mastery. The golf robot can fire shot after shot with stunning accuracy from a long distance. This ad for the European Tour has the added benefit of offering the “coolness” factor to all ages. (Source)

And, naturally, this has all evolved into anything but fun and games. The evolution of PETMAN is an actual Terminator-like robot called ATLAS. Meanwhile, even mainstream institutions like Cambridge and major international human rights groups say that “terminators” are one of the greatest threats to humanity.

As you’ll see in the next video, ATLAS has advanced quite a bit since its introduction in July of 2013. While ATLAS still proceeds slowly on its tether, it is moving with far better agility than its previous incarnation:

And now even better still as it takes to the outdoors. Still tethered, but the plan is for a self-powered version:

However, these are still just workhorses and advanced tools of human mimicry. For these creations to get to the next level, they have to become self-aware and eventually autonomous. The computing power is certainly arriving as evidenced by what has been achieved through flying drone swarms and work being done on autonomous, interconnected battlefield systems such as MUSIC.

Transhumanist icon Ray Kurzweil’s website featured a story about Samsung’s Roboray, which just got a major upgrade. Roboray still has an obviously robotic physical look, but its “intelligence” is increasing:

Roboray can now build real-time 3D visual maps, so he can walk around without being spaced out and wandering off.

Roboray has stereo cameras (one in each eye), allowing him to build a mental map of its surroundings, and to “remember” where he has been before. That kind of autonomous robot navigation will also come in handy in places where there’s no GPS signal.

Furthermore, its walking demo as shown in the video below make it appear more humanoid than its initial outward appearance, something desirable for those who seek to make the human-to-robot emotional connection:

Roboray walks in a more human-like manner by using “dynamic walking” — actually falling at every step, and using gravity to carry himself forward.

Which is how humans naturally walk. So this makes him better accepted by people and more energy-efficient, says team leader Dr. Walterio Mayol-Cuevas, Deputy Director of the Bristol Robotics Lab.

For the next phase of development, it is always important to watch the battlefield, but we should also watch for the signs of marketing that aims to normalize the robot presence in our day-to-day lives.

This is being attempted through next-level artificial intelligence called Embodied General Intelligence (essentially, common sense). The next phase of humanoid robots are those that become virtually indistinguishable from us. There is no lack of effort to try to make this happen. Here is an ongoing Indiogogo project to give humanoid robot Adam Z1 a real brain. This is an effort that also has parallels with NeuroGaming. These two videos leave nothing to the imagination as to what the endgame is: not only the economic acceptance of robots, but the emotional acceptance.

Not that long from now, full-sized humanoid robots will be in wide use as personal assistants, etc. And cheaper versions will be widespread as toys before long: think “RoboSapien with a cute face and a cloud-based mind.”

GENI Lab

So the team has formed GENI Lab, whose medium-term goal is “the creation of a life-sized humanoid robot featuring a realistic, emotional face and personality; a fluidly moving body, based on the integration of analog, digital and mechanical control; an adaptive general intelligence, utilizing the OpenCog architecture and the DeSTIN machine vision system; capability of simple — but socially and emotionally savvy — communication in English and ultimately other natural languages; and communication about its physical environment and its tasks and behaviors therein.”

iCub

This robot is an attempt to put it all together, including tactile sensations and learning from those contact interactions. From Science World Report:

What if a robot could act and even feel like a human? Researchers are getting a bit closer to accomplishing that feat. They’re working on a robot that feels, sees, thinks and learns like us as the lines between man and machine become a bit more blurry.

The robot is called the iCub robot. Currently, twenty European laboratories have an iCub, which was developed in Italy. In the future, researchers hope to develop the robot enough so that it can learn and think like a human does. Although this may seem farfetched, decades of research into cognitive psychology and the human brain have given researchers knowledge about language, memory, motor skills and perception that could help design such a robot.

“The application of cognition in technical systems should also mean that the robot learns from its experiences and the actions it performs,” said Frank van der Velde, one of the researchers, in a news release. “A simple example: a robot that spills too much when pouring a cup of coffee can then learn how it should be done.”

[…]

“The new iCub has a skin and fingers that have a much better sense of touch and can feel strength,” said Van der Velde. “That makes interaction with humans much more natural. We want to ensure that this robot continues to learn and understands how people function. This research ensures, for example, that robots actually gather knowledge by focusing on certain objects or persons. In areas of application like healthcare and nursing, such robots can play an important role. A good example would be that in ten years’ time you see a blind person walking with a robot guide dog.” [emphasis added]

Apparently, even guide dogs are slated to be part of the outsourcing to computers and robots, which threatens to extinguish up to half of all jobs in the near future.

Poppy

Poppy is the latest humanoid robot and presents the first of its kind: Poppy was created by a 3D printer. A group of French researchers were able to cut costs by a third utilizing the latest 3D tech. Poppy’s creators have focused on a biologically inspired walking motion that they hope will allow for better human-to-robot interaction. You can see the incredibly human-like, heel-toe motion in the video below.

To begin with, it has an articulated spine with five motors – almost unheard of in robots of this size, but one of the ongoing topics at INRIA Flowers since its first humanoid, ACROBAN, from several years ago. The spine not only allows Poppy to move more naturally, but helps to balance the robot by adjusting its posture. The added flexibility also helps when physically interacting with the robot, such as guiding it by its hands, which is currently required to help the robot walk. (Source)

Outside of the military, the truly intelligent humanoid robots have only attained a very small physical stature like the several mentioned above. However, a new development by Spanish robotics company, PAL Robotics, is a first step toward offering a human-sized robot that can presumably better fit into its human-occupied surroundings. As you’ll see in the video and diagram below, the REEM-C is quite slow, but does demonstrate some of the fluid movements of other humanoid robots, while measuring in at a respectable 5 ft. 4 inches. This marks another level in the attempt to acclimate humans to the presence of intelligent robots. And the outsourcing continues . . . always accompanied by delightfully creepy upbeat music.

REEM-C

Romeo

a href=”http://projetromeo.com/”>Romeo seeks to become the leader in the areas of robotic care giving and personal assistance with a more emotional element. Romeo builds off of a smaller humanoid robot called NAO that generated more than 5,000 sales or rentals worldwide.

The robot has the size of a child of eight years (1.40 m) and weighs a little more (40 kilos). To be as light as possible, its body is made of carbon fiber and rubber, and was designed to avoid the risk of injury to the person that will attend. Today, Romeo can walk, see the three-dimensional environment, hear and speak, much like NAO

[…]

The timeline for testing the robot in real-world conditions is projected for 2016, with the final objective of being able to use it in a retirement home in the 2017 or 2019.

Funded in part by the French government and the European Commission, the Romeo project budget totals 37 million Euros over the period from 2009-2016. (Source)

NASA’s Valkyrie

NASA has even entered the fray with its entry into the latest DARPA Robotics Challenge December 20th and 21st in search of the $2 million prize. The focus for the challenge is on emergency response where robot entrants will have to perform 8 consecutive challenges. NASA’s robot has been named Valkyrie – from the Old Norse meaning “chooser of the slain.” The legend says that this female figure decides who lives and dies in battle, bringing their chosen to Valhalla. (Source) Nice imagery.

IEEE Spectrum got the first look at the robot recently and shared the details. Valkyrie is 1.9 meters tall, 125 kilograms, and has 44 degrees of freedom. The battery-powered humanoid robot was built by NASA’s Johnson Space Center team in Houston, in partnership with the University of Texas and Texas A&M.

Valkyrie is equipped with a large amount of sensors: cameras and LIDAR in the head, more cameras and sonar in the abdomen, and even more cameras in the forearms, knees, and feet. All of these data won’t be transmitted back to Valkyrie’s operators all at once, but being able to take advantage of whatever sensor is most relevant to the robot’s current task should help Valkyrie be fast and efficient, whether it’s being teleoperated or functioning more autonomously. (Source)

Here is the Valkyrie in action as it preps for the DARPA Robotics Challenge:

Update: NASA’s Valkyrie, despite being called a “Super Hero Robot” by many in the tech press, turned out to be a super failure at the DARPA Robotics Challenge, scoring 0 points. The top award went to Google-owned, SCHAFT, as seen in the video below, which scored 27 out of 32.

One thing that was critically demonstrated at this year’s challenge is the huge gap between what is conceptual and what is practical. None of the robots sufficiently passed the challenge to drive a vehicle, for example. Even the winner, most would agree, came off incredibly clunky, despite some impressive functions such as manipulating small objects like keys in order to open doors.

However, hope seems to be on the horizon for robotics engineers: “super muscle” that would equip robots with 1,000 times the strength of humans. The secret lies in the cells that comprise muscle, as Extreme Tech explains:

Each cell lining the muscle’s length has its own limited contractile ability, one driven by tiny motor and ratchet proteins that crumple up a cell and refuse to let it relax. Rather than being a single continuous muscular action, your bicep contracts as a result of millions of tiny, discreet contractions all working in concert. That being the case, it’s always been something of a misnomer to refer to robot muscles as muscles, per-se — despite being visually similar, they are really just servos, or hydraulic pumps, or whatever else. This week marks the first proof of concept for an attempt to change that, to make usefully strong artificial muscles that actually work like muscles — with all the advantages and disadvantages that come with it.

The research comes from the US Department of Energy’s Berkeley Lab, and it focuses on an amazing material called vanadium dioxide. It’s already used widely in electronics and other industries, mostly because it has an amazing ability to switch between being an electrical insulator and conductor, based purely on temperature.

[…]

The advantages of this model of muscle action are numerous. Think about the fact that, despite being made from relatively weak constituents like protein and fat chains, animal muscle has only now been surpassed by technology. Of course, a bulky, pneumatic Atlas muscle could outperform both this and a human in terms of raw strength, but not while remaining so light and compact, nor is it so instantaneously responsive as a system based on micro-meter movements. The actual energy costs per unit strength are unknown right now, but has the potential to be extremely efficient, relative to current solutions. (Source)

You can see a short video description of this polymer technology here.

iCub – The robot that has a sense of self.

It appears that all of the components are coming together which are leading to the possibility of a true terminator robot. And once these humanoid robots merge with fully autonomous intelligence, we will have witnessed the next stage of evolution toward the full Transhumanist dream of The Singularity – the moment when computer intelligence surpasses that of humans to such an extent that humans become practically redundant.

Robots already have transformed our human world and are rapidly evolving. If The Singularity is reached, in tandem with military funding and direction, we can expect the darker version of science fiction to rise above any notion of attaining human freedom and leisure on the backs of our machine counterparts.

I find it ironic that these sentient robots are only made so by injecting them with humanity. But we are continuously bombarded by the global elite with the message that humanity is the core problem. The fact is that robots are nothing without the boundless potential that resides within the human brain; nothing but a computer doing fancy tricks that imitates us. True, we have a long way to go to reach our full potential and mitigate our self-destructive tendencies, but a complete replacement of our species at this juncture appears to be short-sighted and is obviously artificial.

How much of this technology should be embraced, and how much rejected? Now is the time to set the boundaries. Please leave your comments below.

Image Credit

Sources:
http://www.2045.com
http://www.kurzweilai.net/humanoid-robot-builds-real-time-3d-visual-maps-walks-like-humans

Timeline of humanoid robotics:
https://en.wikipedia.org/wiki/Humanoid_robot

Updated 8/17/2015

Read the Latest: ATLAS: Next Generation of DARPA Humanoid Robot Released