Transferring the principles of Industry 4.0 beyond the industry itself was one of the ideas of FME student Martin Juříček. He designed and put into operation a semi-autonomous system in which a robot replaced a paramedic in collecting an antigen test sample from the nostril. With an idea on how to transfer the principles of automation and robotization to the healthcare sector, his work ranked among the TOP 9 dissertations on Industry 4.0 in the Siemens Award 2022.
"I finished my bachelor's degree during the first wave of the Covid pandemic. Exactly for this reason I got the idea that my diploma thesis could focus on some solution in this area," recalls Juříček. After consultation with his supervisor Roman Parák, the choice was made to build a robotic platform for antigen testing, using a number of technologies typical for Industry 4.0. "We wanted to show that robotics does not belong only to the area of the classical industry, but that it is very extensive, and medicine, for example, is one of the areas where further development can occur," explains Roman Parák from the Institute of Automation and Informatics (the openTube robot-lab technician has already been created at the same institute, and we have already covered this here in the past).
The system cost Martin Juříček two years of hard work and he devoted himself to this development mainly in his free time, in addition to his other school duties. Gradually, he had to study technologies with which he had not come into contact before. "These included, for example, a 3D camera, a silo-torque sensor, or the use of an end effector, a device with which a robot grabs an object. But it was not only the hardware, but also a number of software solutions that were completely new to me," explains Juříček.
The aim was to mimic the real testing as much as possible. "The system can detect a person's height and identify whether the patient is standing in the camera area. Using a 3D camera and image processing, we capture the face and the robot receives information about the 3D spatial point in the center of the nostril. Following the generated trajectory, the robot then performs sampling in a circular motion at the bottom of the nostril," Juříček describes.
Control is also part of the system. "At the beginning, the laboratory technician registers the patient in the information system. In addition to the basic data and a specific identification number, the registration also contains visual identification. Before the collection itself, the system uses neural networks for facial recognition to validate whether the right patient is standing in front of the robot to avoid mistaking," adds Parák.
It is clear from the video that it takes a certain amount of courage not to dodge the robotic medic. So how did the "training" of the test robot take place? "My first tests were shy and were performed with a great distance," Juříček confirms with a laugh and adds: "The collaborative robot was equipped with a very sensitive force-torque sensor. I knew that if I didn't like anything, I could just stroke the robot lightly to stop it, so it was safe."
Juříček is still studying at the Faculty of Mechanical Engineering and is doing his doctorate at the Institute of Automation and Informatics. At the same time, he manages to apply his extensive experience with robotization in a completely different field than medicine: agricultural technology. "I work at FRAVEBOT, where we are developing a similar system, but designed to pick tomatoes or strawberries. I draw a lot of experience from my diploma thesis, even though I am putting it into practice in a completely different field," says Juříček.
For his diploma thesis, which he submitted last spring, he recently won 4th place in one of the Siemens Award categories. According to the leader, it was well-deserved. "It could easily have been split in two or three separate theses. But he managed to achieve all this in a single brilliant work," Parák praises his thesis writer.