Robots_Fraunhofer_IZM

ROBOTICS 101: HOW SMART MACHINES ARE CONQUERING OUR WORLD

In 2021 alone, robots are expected to generate a turnover of over 41 billion US dollars. That is equivalent to the GDP of a medium-sized European economy. Time to take a look at robotics as a scientific discipline: What makes them different? What is driving the current market, and how does Fraunhofer IZM contribute to it?

Let’s start with the brass tacks: What is a robot?

One established definition comes from the International Federation of Robotics: “A Robot is an actuated mechanism programmable in two or more axes with a degree of autonomy, moving within its environment, to perform intended tasks. Autonomy in this context means the ability to perform intended tasks based on current state and sensing, without human intervention.”

This defines several aspects that make a machine a robot. For example, it separates robots from remotely operated vehicles, like the ones that are used in Fukushima to clean up the rubble. Those lack the required degree of autonomy, as they are merely executing the orders of their operators. The definition also has robots working on axes, which is something that makes sense when you speak about industrial robots but, less so when we talk about service robots. Service robots have come to represent a significant share of the market, a share that is expected to overcome that of industrial robots within the next few years. So, we have a general definition of robots that excludes a significant part of what we would nowadays regard as robots. Why is that? It is because robotics has undergone an evolutionary leap in the last decade: Robots in general and service robots in particular are slowly making their way into the consumer market.

What skills do they have?

Technologically speaking, robots have many types of properties to consider: dependability, configurability, adaptability, mobility, the ability to take autonomous decisions, cognitive capacities, and the ability to perceive, manipulate, and interact with the world around them. It is the sum of all these that makes for a functioning robot – no matter if it is an industrial robot or a service robot. But it is also the challenges relating to these technical properties that have been keeping service robots from entering our day-to-day life so far. It is one thing to build a smart robot that is as big as a truck and immobile, but something else to fit an electronic brain into a case that is as small as a shoe box and can move around your living room.

Let us take a look at one example: Take a bionic insect, a robotic ant. Its sensorics components are made up of a stereo camera to self-locate and to perceive its surroundings as well as optical sensor chips to detect structures on the ground by which it measures the travelled distance. To move said distance, our robot insect uses piezoceramic legs. To manipulate objects, it has piezoceramic mandibles. It communicates with other bionic ants via a radio module and a small antenna on its back, and it ”thinks” with an embedded AI chip. All of that is powered by a power supply unit that is attached to the back of the ant and converts the 8.4 volt from the LiPo cell batteries to the 300 volt that are required for moving the mandibles and the legs. The batteries are recharged by wires on the front of the ant, through which it can connect to a docking station. Together with all the wiring on the ant itself, you have quite a bit of machinery that needs to find space on a very small frame.

The dawn of the robots: Why is it happening now?

Luckily, robotics have benefited from advances in many related fields. Smartphones, the automotive industry and communication technologies have contributed tremendously to the development of new robots; be it with better sensors, greater computing power, or improved connectivity.

Robots have a rich and storied history. While it all started out as an attempt to build metal men, it quickly shifted to machines to do labor for us, because the technology to build robots that are smart and adaptable enough to function independently in a world as complex as ours simply did not exist. Robots that are highly skilled at doing only a few specific tasks and nothing else – that was achievable. They might not look like humans, but they massively outperform them at predetermined (albeit rather simple and repetitive) tasks.

There have been attempts to overcome this complexity hurdle by putting robots alongside humans: collaborative robots. Whenever a situation came up where the robot was at a loss, a human operator could step in and perform the action that was beyond the robot’s capabilities. Once that was done, the robot returned to its regular workflow. The issue was that these robots had to have so many safety precautions built into them to avoid endangering the humans working alongside them, that the entire concept was simply not worth it. Eventually, the focus shifted back to robots that worked independently, but on a rather limited set of tasks and under rigid circumstances.

Leaving the industrial niche

All the while, the technology behind robotics kept getting better and better. Sensors got more accurate and perceptive, AI got smarter, form factors got smaller, and robots eventually went from doing simple tasks with limited automation to more complex work and a higher degree of automation – the advent of the service robot.

In the recent years they have leapt the technological barrier that kept them from entering the consumer market. Nowadays (service) robots have model-based spatial awareness and independent power supplies, can communicate with their surroundings, detect and execute tasks intelligently, and even recognize human behavior. Or in other words: They are small enough to function in a human context, smart enough to interact meaningfully with it, and powerful enough to provide real value.

Three examples of service robots that are already seeing widespread use are robotic vacuuming cleaners, search and rescue robots, and maintenance robots. All of these are navigating autonomously and work without human support, they have their own power supply, and they follow complex routines (vacuuming cleaners maybe a bit less so) to identify and solve problems. In other words: technology has finally caught up. Service robots have arrived. And with the successful leap into the consumer market comes the money and attention of big companies and their research and development departments.

Industrial robots are well established as the backbone of many branches of the economy. Be it in manufacturing, healthcare, agriculture, or logistics: in some of these fields, they have been used for as long as 60 years, whereas service robots have only now been starting to see widespread acceptance. Yet, the annual turnover for service robots is expected to reach 19 billion dollars already in 2021. That is merely 3 billion dollars less than with industrial robots, which had a lot more time to carve out a market for themselves. If we look at the growth rates, it becomes even more apparent: the market for industrial robots is expected to grow by 12% in 2021 – a respectable amount in itself. But service robots could see an increase of up to 23%. An impressive number, given the fact that they are already almost on par with their industrial brothers.

So, what does the future hold for robotics?

The market is diverse and its players legion. Industrial robots will get smarter and more powerful, but most likely without any big leaps in terms of functionality. Manufacturing, agriculture, and logistics robots will become faster and smaller, but in the end, they will still do the same tasks. One development that we might see is simply an increase in numbers, which might eventually replace parts of the human workforce. Already we have fully automated warehouses where only a few humans are required to keep everything operational. With more sophisticated technology and wider applications, this trend will spread into more and more disciplines, reducing the need for manual human labor even more.

It gets more interesting once we look at service robots. A sector that is comparatively young and as such still holds a lot of untapped potential is the medical domain. Robots that are small enough to be swallowed or injected could change the way we treat ailments. Surgical robots are already well-established and researched, but when we reach form factors of only a few centimeters to millimeters, there is a lot of uncharted territory. Questions regarding the power supply, autonomous movement (in contrast to being moved by a force in their environment), communication, sensors, and possible treatment applications are seeing major research efforts. Each factor has essentially already been mastered separately, but to build a prototype that combines all of them into a fully operational unit is the next big challenge.

The same holds true for robots that are expected to work independently in hostile environments. Once the robot has entered these environments, it is very difficult to physically interact with it, and the robot is mostly on its own. The necessary degree of independence and reliability is immense, but once that threshold has been surpassed, these kinds of robots possess lots of potential. Be it in the deep sea, in outer space, or anywhere else: Robots that can work independently for a sustained amount of time can solve some of humanity’s most pressing issues. They could clean the oceans, collect space debris, or dig up unexploded ordnance.

A third example, and maybe the most intriguing pathway, is the social robot. With the technology to produce robots that can move like humans, it is only a question of time until robots also act like humans. You can already buy robot cats and dolls that actively engage in social interaction with their owners, adapt to their owner’s behavior, and try to be as lifelike as possible. With the technological hurdles slowly disappearing, it will be AI that dictates what is possible and what is not. And the possibilities are numerous.

The bottom line is: Robots are extremely versatile. Flying, walking, swimming, or even caring – given enough time, anything is possible. To the extent that the important question is not anymore: “What can we do?”, but rather: “What should we do?”

In psychology, there is something called the uncanny valley. It describes a dip in the graph that correlates a robot’s human likeliness with its likability. Or in other words: How well does a robot resemble a human being, and how much do humans sympathize with it. Research has shown that there is a certain threshold after which an increase in likeliness results in a decrease in likability. Robots that look like bulky metal men are cute, but robots that almost look like real humans (but not quite) are scary. It will be interesting to see what will happen once technology allows us to leave the uncanny valley.

Where does Fraunhofer IZM come into play with robots?

Our top priorities for the next five to ten years are in fabrication schemes, power and energy, robot swarms, navigation and exploration, and brain-computer interfaces. Additional focus is being put into biohybrids and bioinspired robots, AI for robotics, and medical robotics.

We are doing a lot of fundamental research on sensors, actuator engineering, communications, AI, and power supplies. With funding from the EU, which is investing heavily into robotics and specifically into the topics perception, cognition, and navigation, we are researching new materials and manufacturing processes, structurally integrated electronics, power distribution, miniaturization, and biohybrids. Some of our applied research has been on projects with exoskeletons for paraplegics, smart membranes for underwater robots, or electromechanical vests that support their wearers with strenuous tasks. With exoskeletons, the specific challenges come from integrating soft actuators (like pneumatic rubber tube actuators) into textiles, mobile energy supplies, and embedded control systems that respond intelligently to constantly changing parameters. Autonomous underwater robots on the other hand require specialized energy supplies, sensors that work in liquids and low light environments, navigational tools that work with the higher degrees of freedom of the underwater landscape, and different materials in general. This can enable robots to perform tasks far beyond the capabilities of humans, e.g. measuring underwater structures or inspecting pipelines. The challenges are diverse, and almost every aspect of our lives could be changed with the introduction of robots. The only question is how.

Let’s find out.


IT CAN BE ALSO INTERESTING FOR YOU:

Leave a Reply

Your email address will not be published.