“In the context of the 10th anniversary of the first Athens Science Festival, the Hellenic Industrial Property Organisation supported the unique celebration of science and innovation on April 21, 2024, at Technopolis, with participation in a dialogue panel. The Organisation was represented by Mr. Demetris Stafylas, Electrical Engineer & Computer Scientist, Head of Search Directorate, who participated in a discussion panel at Innovathens titled ‘Humanity Through Robotics’.”
With journalist Giannis Rizopoulos as the moderator, curator of the Athens Science Festival, the topics discussed included robotics and automation technologies, which are increasingly becoming part of our everyday lives.
The panel consisted of Mr. Evangelos Papadopoulos, Professor at the School of Mechanical Engineering, National Technical University of Athens, Mr. Konstantinos Nestorakis, Innovation Engineer, IT Innovation Center Group OTE, Robotics, and Mr. Giannis Kounas, General Manager at ASBIS Hellas Single Member SA.
Mr. Stafylas, as part of his participation in the panel, presented the following:
Presentation of two case studies:
- The award-winning Spanish engineer Elena Garcia Armada, named the best inventor of 2022 through an online public vote, for her invention of a robotic exoskeleton suitable for children with disabilities, featuring joints with controlled stiffness through the use of artificial intelligence algorithms (a combination of robotics innovation and healthcare technologies / EP3225363B1, EP3009240B1).
- The invention of the Norwegian company Stingray Marine Solutions, which was a finalist for the best invention of 2019 in the small and medium enterprises category. The proposed invention involved a robotic submarine equipment within a salmon fish farming unit where, through artificial intelligence image processing algorithms, it was determined whether the salmon was infected by harmful sea lice, and in case of infection, a mechanism for green laser pulse irradiation was activated to eliminate the parasite (a combination of robotics innovation and marine engineering / EP2531022B1).
Next, Mr. Stafylas explained the operation of autonomous robotic mechanisms, which is based on interpreting the environment in which they are placed by collecting multiple data through their sensory systems, processing them in their embedded processors using advanced Artificial Intelligence/Machine Learning algorithms, to adjust their movements and achieve the set objectives with greater precision.
For a robotic mechanism to interact effectively with its environment and operate autonomously, it is crucial to further develop its robotic senses (visual, auditory, tactile, olfactory, and gustatory). Continuous advancements in sensor technology, computer science, and digital communication systems make it possible for robotic mechanisms to be applied in an increasing number of technical fields.
The use of the ChatGPT model in robotics was mentioned, aiming to further develop robotic intelligence. The 7 distinct types of robotic intelligence were discussed in detail (linguistic, logical-mathematical, spatial, bodily-kinesthetic, creative, interpersonal, and naturalistic), where the GPT model is expected to be applied.
There was also a concern raised about who will have control over the operation and responsibility for the actions of autonomous robotic mechanisms, as current human-centered legislation focuses on ethical and natural rights of humans. In a scenario where robots are accepted as entities with perceptual capabilities, a new legal framework may be needed to ensure their existence’s protection.
Mr. Stafylas emphasized that the decision-making process of AI models requires transparency, where every citizen has the right to an explanation (explainable AI) on how a decision was made during data processing; otherwise, we may face the risk of algocracy (the tyranny of algorithms).
In conclusion, he mentioned that Artificial Intelligence models rely on exploratory data analysis, thus a regulatory framework is required to protect data ethics, ensuring the ethical use of data to train these models. Otherwise, any incoming biased data in favor or against a social group would result in the reproduction of this bias in AI decision-making models.”
