Robots in Healthcare: Challenges of Integration with Karol Miller

For today’s episode, we have  Professor Karol Miller with us. He is a specialist in the area of modeling and computer simulation in biomechanics. He is the Winthrop Professor of Applied Mechanics at the University of Western Australia, the Deputy Head of the School of Mechanical and Chemical Engineering, Director of the Intelligent Systems for Medicine Lab and Director of the Computational Geomechanics Laboratory.

Transcript:

Elisa Muñoz: Welcome to one more episode of Builder Nation. For today's session we have professor Karol Miller all the way from Australia. He's a specialist in the area of modeling and computer simulation in biomechanics.

Karol Miller: Thank you. Thank you so much for having me. I am Karol Miller, born and educated in Poland. I am an aerospace engineer, but for most of my career, I worked in biomechanics and robotics. So after completing my PhD in Poland, I went to work in Japan for government labs day. After a few years in Japan, I moved to Australia,  I've been in Paris, and now, I have been at the University of Western Australia for 26 years.

Elisa Muñoz: Now I'm really curious. I mean, you're an engineer, right? At what point of your career you were like “Okay, I'm going to start working with robots”?

Karol Miller:
So, my master's was in biomechanics, then I did a PhD in robotics and a doctorate of science, again in biomechanics. And wherever you look, wherever you turn your head, now you have essentially mechatronic systems around you, right? So there is a car, a fridge, a phone, everything is a mechatronic system combination of mechanical engineering, electrical engineering, and computer engineering or computer science. Right?

Elisa Muñoz: How can robotics be applied to the surgery or to the actual research?

Karol Miller:  Robotics has been around for a very long time. And I, when I first attended my robotics conference, I was a PhD student at the time in Poland. And I think this conference was in 1992 called “The romance robotics and manipulator systems”. One of the prestigious conferences in the field.

I had the keynote from a very famous American robotics researcher in which he proclaimed that within 10 years, every Japanese household would have a domestic robot in their apartment. And every second American household would have an autonomous domestic robot and that didn't happen. And five years later, I had the same prognosis from a different keynote speaker and 10 years later, again and again. So we need to ask ourselves the question, why are we not surrounded by robots? Look through your window. Do you see any robots? No.

So, we put the blame on two computer scientists who promised artificial intelligence and, and having to deliver it, right? So, computers or computer programs, computer based control systems seem to have significant difficulty with uncertainty.

Elisa Muñoz: But what happens when it comes to actually doing a surgery with AI?

Karol Miller: So I think that the idea that we can have an ultra novel century, it can safely be abandoned. I am very radical, so the robot can be used as a tool in the hands of a human, like, for example, the Vinci robot. It's called the robot, but it's not really a robot. It is just a, what we call a copy of my plate. It copies the movements of a surgeon at the very small scale inside the body of a patient. Right? So it does not do anything on its own. It just translates the motion and all thinking and decisions of a surgeon to precise manipulation at, at very smaller landscapes. Right? So this is of course, very good.

And then I'd also like to point out that there is very significant resistance to both, or say, psychological and social aspects of the situation, both the medical profession and the patients are not happy or ready to delegate decision-making to a machine. Right. So, no, that will not happen anytime soon, apart from perhaps some extreme environments, like for example, a century on the space station, where way may be don't have access to a human experience, or maybe maybe on battlefield, but even in on battlefield, I still believe that there was a better prospect for tele operation, essentially, that somebody is doing something.

And the actions of a doctor at the base are copied by a machine on the ground rather than a machine doing something autonomously or on its own.

Elisa Muñoz: Yeah, totally. I mean, I think that we, as society, might not be totally prepared in order to replace it if we can put it like that, like human interaction or human decisions for robots you know?. And how precise are they? You know, when it comes to the movements, when it comes to translating?

icon-spotify-white

Karol Miller: The mechanical engineers have done an excellent job. So very, very precise and much, much more precise that can ever be achieved by a human. Just imagine microsurgery. We are talking about very precise motions by a machine in very reliable methods. So this is all under control. This is not even robotics. This is very clever mechanical engineering.

Elisa Muñoz: Let's say that their robots are ready to receive the information and have already processed all the information from the scans, how is that presented to the surgeon?

Karol Miller:  A very important step is to actually not show anything to the surgeon when the image is considered normal. That's the main thing.

Because out of these 1 billion say x-rays that it will be 990 million probably showing nothing to worry about.

So, now the ones which have something to worry about are automatically normally labeled by the system. Then the machine says, “I suspect tuberculosis, or look, I suspect lung cancer. I am still using that example of, of, of a chest x-ray.” So, in my opinion, by far, the most important thing is to, to essentially take the vast majority of images labeled as normal and not show them until the end.  

Elisa Muñoz: And I mean, since we're talking about future improvements, do you think that, let's say in 30 years we will have life filled with robots?, You know? Like that every time that we get injured or whenever we go to the hospital, we will have a robot in there?

Karol Miller: So this is a very deep or high-level question. And there are a number of people, famous scientists who put forward in my view, a very convincing argument that it is in principle, impossible, right?

Which means 30 days or 300 years or 3000 years, it makes no difference. It is in principle impossible. If you believe in, in that so-called Gabelli and argument of Roger Penrose. And I actually believe that this argument is correct. And I have seen that I have worked in robotics for 30 years, and I have seen zero progress in, in this ability of robots, to run the streets.

I have seen enormous progress in engineering development for industrial robots or for image recognition or for localization and mapping and things like this. But for example, self-driving cars, I don't believe, I don't think this can happen.

It can happen in China where they will structure the road, right? The road will be prepared for four cars, but just self-driving on random roads in Australia Outback, no. I don't believe in such things. It's not because I am thinking that technology is not mature enough, but I believe that there is in principle a difficult day, essentially that technically speaking that many operations we conduct in our long computable, and of course, non computable operations can not be replicated on a computer because it computes.

Elisa Muñoz: And any last advice that you can give to our audience.

Karol Miller: It's important to avoid working hard because if you work hard, there is little time left for thinking like on tangents and, and good ideas usually come in unexpected times when you are on the beach or maybe in the dream, or who knows. Well, so my advice is to avoid working very long hours, because then there is no space for these thinking processes.

Elisa Muñoz: That's the best advice I've ever heard. Thank you. And thank you so much for taking the time to be here. We really enjoyed hearing your experience.

Karol Miller: Thank you so much. It was a very pleasurable experience.

This interview was brought to you by ControlHub, the most intuitive purchasing software to request, approve and track all your business purchases.
Thank you! You are now a member of Builder Nation
Oops! Something went wrong while submitting the form.