What do we actually want from bots?

As an anthropologist, Massimiliano Simmons observes a number of robotics in Darmstadt to see what robots can tell us about humans.

“The experiment will take about five hours,” Argang told me. Arjang Ahmadi is a student at TU Darmstadt. Together with his fellow student, he is busy preparing measuring equipment. Passive sensors are placed all over my body, which reflect infrared light. A series of EMG sensors are placed on my right leg to measure the electrical activity of the muscle nerves.

We started at 8am, but it took a long time before the experience really started. Infrared cameras have not yet been calibrated and the application of sensors to the body is not intuitive, as each body is different. But what frustrates them is for me a goldmine of information.

A series of sensors measure the activity of the muscles in my right leg and transmit it to a computer.

For as they prepare my watch, I watch them. I feel like one of the guinea pigs in Douglas Adams The Hitchhiker’s Guide to the Galaxy. Science fiction writers tell us how mice are subject to our human experiments only voluntarily because they are in fact extraterrestrials secretly studying the behavior of scientists.

Likewise, as an anthropologist, I am in Darmstadt to observe scientists. I want to understand contemporary “technical sciences”. Technological science is characterized by practices that rely heavily on the use of technology to produce and study phenomena. In this case, it concerns robots used to study human movement.

symbiotic model

Everyone knows videos of robots failing to perform seemingly simple actions, such as opening a door. It’s the most popular form of robotics research: you’re building a robot that will take on part of the human tasks.

But in Darmstadt they do it differently. Professor Andre Seyfarth tells me, who Lovelabour Leads to where my experience occurred. Despite being a professor of sports sciences, he sees his research on human movement in light of larger philosophical questions about human nature.

See also  James Webb Space Telescope reveals solar shields and continues its journey steadily | Science

He sees his work as an alternative model for looking at robots. While the American model aims to replace or surpass humans, in Darmstadt it is essentially a symbiotic model: humans and machines work together. Every research into robotics aims to tell us something about humans.

While the American model wants to replace the human, it is a symbiotic model in Darmstadt: human and machine work together

According to Seifarth, this choice is not independent of institutional choices. American research is often sponsored by the military, while funding in Germany comes from German Research Foundation (DFG). It is no coincidence that in the United States they see robots primarily as supernatural beings. According to Seifarth, there is more room for basic research in the German system.

So what we want to do with robots is not a foregone conclusion. It depends on societal choices about what kind of research we want to support and what we want to invest money in.

bullying people

At first glance, most experiences come in Lovelabour About tricky ways to bully people. For example, test subjects must walk on an elevated platform where the height of this platform changes unexpectedly. Next, the researchers study how the test subject does everything they can to avoid falls.

A similar project, in collaboration with Dutch colleagues, made the test subject “Backpack”. Here, too, the backpack can unexpectedly push a person to the left or right while walking.

On this platform, test subjects must move while allowing the examiner to change height unexpectedly.

But all of these experiments are based on the above philosophy: they involve the study of human motion in order to make better robots, but robots that in turn will serve the human body.

The experience I am sharing is a step in that direction. I have to walk on a treadmill, either without an “exoskeleton” or an “exoskeleton”: a suit that can be worn around my legs that makes walking easier. What I wear is a passive exoskeleton, because it doesn’t have a drive like an active exoskeleton. Instead, it works with a series of walking aids and springs.

In this outfit, I have to walk countless sessions on the treadmill with different exoskeletons.

The aim is to see what form such an exoskeleton takes to increase efficiency in the form of reduced muscle and energy consumption. Hence the need for all the sensors and cameras that closely monitor how my gait changes in the different variations of the experience.

Android as a mirror

At first glance, the goal appears to be better robotic prosthetics, but the reality is more complex. Technologists usually do not distinguish between practical and theoretical goals. Yes, they want to make usable prosthetics here, but at the same time they are using robots to better understand human movement.

But why should human movement be studied by robots? Why is it not enough to study human bodies? When I asked this question to them, they got different answers.

Technologists don’t really distinguish between practical and theoretical goals

First of all, there is simplicity: the human body is complex. Let the human walk and you will have everything on your plate at once. Replace that human with a simple robot and you have a concept model, with just a few joints and muscles to consider.

In addition, dealing with bots is easier. You can’t just cut a muscle in the human body for the sake of experiment, due to practical and ethical objections. This is possible with bots.

A robotic leg that robot scientists use to mimic and study human movement.

Sefarth gives two other reasons. For him, the robot is also a judge. Your theoretical model can be put together quite well, but the bot gives the model the ultimate test: does it also work in reality? He also sees robots as mirrors: they show what we understand about human movement. If we can’t replicate it, we don’t understand it yet.

Go to robots

Some experiences in Lovelabour It also focuses on human-machine interaction. Seyfarth gives an example of two robots that should see together. People can easily coordinate such work. But how do they do that?

This question becomes more important in the future as collaboration with robots will become the norm. What if your colleague on the site is a bot? But the same is actually true of exoskeletons. This is also basically a human-machine collaboration.

Any social coordination assumes a bond of trust. The same goes for the exoskeleton: when do people trust the exoskeleton enough to rely on it?

This raises not only technical, but also social issues. Seyfarth gives a simple example: Let’s say someone wants to help you carry your bags. When do you trust this person? Any social coordination assumes a bond of trust. The same goes for the exoskeleton: when do people trust the exoskeleton enough to rely on it? When do people trust their bodies?

In this sense, experimenting with robots tests our social future: How will we deal with robots, once they appear in mass on our streets? This future is not yet fixed, but it depends on our choices. With policy considerations and research funding, the question always arises: What future of robotics do we actually want?

Leave a Reply

Your email address will not be published.