With the announcement of Tesla bot, prototype of humanoid robot, Elon Musk, a technology revolutionary to many, a charlatan to others, in a round of dances has redesigned the boundaries of our knowledge in the field of robotics for the near future, surprising everyone: those who design them, those who study them, but also those who simply wait for them for a long time.
With the announcement from the stage of Tesla’s AI Day, that time magically contracted because in a year or so we should see Optimus in action, a robot that aims to replace humans in the dangerous, repetitive or boring tasks of everyday life. We know its anatomy and operating characteristics but we do not know how and if the experts involved will be able, in such a short time, to cope with still unresolved problems in laboratories all over the world such as hand grip, with respect to which a “human-level hand” is envisaged.
Today, robotic hands are unable to take objects autonomously, firmly, and sturdy. Let’s imagine this environment: a pocket emptier in which there are several objects close together. The hand should first be able to push other objects away to avoid colliding with them and then grab the chosen object with only two fingers because perhaps in its entirety it would not be able to enter. For the current state of research, this simple operation represents a sum of complexities that refer not only to a question of mechanical design but also of software: understanding the environment and using the most suitable types of manipulation to be implemented. Robotic hands have shown great progress, but they have not yet reached the level of dexterity of the human hand. Ditto for walking in unstructured environments, i.e. with physical and geometric characteristics not known a priori. Sometimes spectacular prototypes, such as the bipedal robot Atlas della Boston Dynamics, which we have seen struggling with a parkour session, are and remain platforms on which engineers do research and development. The aim of Bd’s more than thirty-year research is not to create superhuman robots, but to develop the best control and perception systems with the aim of producing robotic solutions to be marketed to assist humans in various activities. An example is Stretch, one of the last born. A robot on an omnidirectional basis that allows you to move between loading docks, travel along the aisles of a warehouse, but also to stop in the load compartment of a truck to unload boxes.
Although we recognize the importance of the capital of technical knowledge that comes from Tesla cars (“our cars are semi-sentient robots on wheels”), we do not know if this know-how will be sufficient to guarantee what is promised if we also think of other giants Japanese like Honda, Sony or Toyota who have already developed robots after decades of development, some of which they have already been commercialized, but always for limited and controlled tasks. We will wait until 2022 to understand if the Tesla humanoid prototype will be downsized or not compared to the initial announcement, which in any case invites us to reflect on the potential that exists when multiple factors come into play for the evolution of a technological system. If in the name of a project they are generated movements of enormous capital, a value chain for robotics, like the one that can be created with the automotive sector, and guiding the process is the visionary ability of a tech guru, so I believe that a technological acceleration can really be achieved that in the long term will lead to intelligent robots that can be marketed at a price reasonable. Just so robotics will earn the title of disruptive innovation.
The American tycoon has decided to invest in robotics now that the Korean company Hyundai Motors recently acquired 80% of Boston Dynamics from Japan’s Softbank, a leading company in the field, suggesting that with Tesla it also wants to outline, on an all-American front, a strong presence in robotics. To do this, it allegedly points to the funding that Darpa, the government agency of the US Department of Defense in charge of developing new technologies for military use, has for robotics.
Elon Musk will be a guest of Italian Tech Week, the event scheduled for 23 and 24 September in Turin. We can hope that his talk will produce such an echo as to draw attention to the topic of robotics which seems to have been forgotten by our politics. As coordinator of the working group that drafted the document of the National Research Program, we were pleased to have identified a scenario in which robotics (considered together with artificial intelligence and others among the strategic areas for the seven-year period 2021-2027 in the industry and aerospace cluster) could have a major impact for large national research areas and innovation, including digitalisation, health, critical infrastructure and clean energy.
In the next National Recovery and Resilience Plan there is no longer any trace of robotics; moreover, artificial intelligence, which instead is present in the document and with which robotics has a close link, is no longer, as in the previous version, among the enabling technologies considered vital for the future of the country. As an exponent of the robotics community, I cannot fail to be struck by the failure to recognize robotics as a lever to focus on to unlock new potential, given the fact that Italy has a tradition in the field which places it at the highest levels on the international scene.
On the other hand, there is no shortage of promising scenarios such as the birth of RoboIT, the first national pole for the technology transfer of robotics, created to enhance the results of Italian scientific and technological research through the creation of startups conceived in the laboratories of universities and research centers of excellence. RoboIT is the result of the collaboration with the Italian Institute of Technology and involves together with the University of Naples Federico II, the Scuola Superiore Sant’Anna of Pisa and the University of Verona. With 40 million euros, Cdp Venture Capital Sgr invests in the future of robotics through the Technology Transfer Fund and in the future it will finance the entire technology transfer chain through the creation of national poles distributed throughout the territory. There Federico II’s partnership with RoboIT testifies to a long history of success of the Neapolitan school of robotics which over the years has led to funding in the context of European projects and international recognition in various sectors of application.
In front of the more innovative scenarios that are presented to us, the question we would like to ask Musk is whether the company actually needs a humanoid or whether it is desirable, more likely, that specific applications can turn into robotic technologies useful in different sectors, from industry to services. In our workshops, this is the direction, the rest seems to be one story among many.