Italian Researchers Unveil Humanoid Robot With Jet Engine
As robots increasingly move beyond the controlled environment of research labs into the chaos of real-world architectural infrastructure, getting from point A to point B has become a serious challenge, such as climbing a staircase.
In response, roboticists have developed a number of solutions: mounting rotors so a robot can overcome obstacles by helicopter, or, in the case of Boston Dynamics, performing backflips that would make the best gymnasts envious.
And then there is Daniele Pucci, head of the Artificial and Mechanical Intelligence Laboratory at the Italian Institute of Technology, who took a bold step by strapping a fully functional jet pack, similar to the one developed by Richard Browning, onto the back of the iRonCub synthetic humanoid, in the hope that it will eventually take off into the sky.
iRonCub looks like a robotic schoolboy equipped with four jet engines that give it the ability to fly like Iron Man. We can at least hope so, since Tony Stark flies with ease, but in reality this is a complex, sometimes fiery and slightly explosive problem, especially for a humanoid robot that was never designed for such things.
Scientists at the Italian Institute are pursuing research in several directions to bring aerial humanoid robotics to the iCub humanoid robot: they are studying flight control, conducting experiments with jet turbines, and working with computational fluid dynamics for aerodynamic modeling.
Some may think humanity learned its lesson about the dangers of creating aerial humanoid robots after first watching “Avengers: Age of Ultron,” but Pucci’s team believes such systems could one day be the first to respond to some 300 natural disasters that kill around 90,000 people worldwide each year.
Looking at individual cases, the balance of natural disasters can be even more alarming. The 2004 earthquake and tsunami in the Indian Ocean claimed the lives of about 230,000 people in 14 countries, caused 140,000 injuries and, as a result, 1.74 million people were forced to take measures and leave their homes.
We have seen many disaster response robots, some humanoid and some not, emerging from labs for more than a decade, often with varying degrees of success.
When it comes to disaster response, humanoid robots have an advantage over more esoteric designs and traditional UAVs, because they can more easily navigate a world already built for humans.
However, when a disaster strikes, much of this human-oriented infrastructure can be damaged or otherwise made impassable, negating many of the original advantages of a humanoid robot. But by combining a humanoid design with the ability to fly, Pucci’s team can take the best aspects of both technologies.
In 2019, a team of IIT researchers reported: “Aerial Humanoid Robotics combines aerial manipulation and humanoid robotics. Thus, aerial humanoid robots overcome the lack of ground mobility with aerial manipulators and extend the locomotion capabilities of humanoid robots to the case of flight. Aerial humanoid robots can walk, fly, manipulate and transport objects, thus offering energy-efficient solutions for payload transport and object manipulation.”
“Aerial humanoid robotics extends aerial motion control to a more reliable and energy-efficient level. In fact, aerial manipulation is often demonstrated with quadcopters equipped with a robotic arm. These robots cannot move using contact forces with the environment, and they often struggle with flying in windy conditions while manipulating an object, requiring precise position control to carry out manipulation tasks. Thus, an additional arm on a flying humanoid robot could establish a contact point between the robot and its environment, thereby making the robot’s position control simpler and more reliable”
In a comment to IEEE Spectrum, Pucci said: “I sincerely believe that aerial humanoid robotics could be used as a testbed for future flying exoskeletons for humans. Richard Browning’s recent successful feat shows the technical feasibility of these futuristic powered exoskeletons. However, we still have a long way to go, and we can use flying humanoid robots to speed up this journey and avoid many human trials.”
According to him, there are also scientific benefits to studying aerial humanoid robots. He believes that controlling a flying humanoid robot raises a number of theoretical and practical questions. For example, a general control framework covering manipulation, contact-based locomotion and flight is still missing, and the role of auxiliary (jet) thrust during contact-based locomotion of humanoid robots is unclear.
Among the first questions are: “At what walking speed is it more energy-efficient to engage auxiliary thrust?” and “How do we handle landing impacts to transition smoothly from flight to walking?”
