There is no doubt that robotics, as a disruptive technology, will change life as we know it over the next 50 years, enriching and augmenting all the aspects of life. The main robotics challenges in the next few decades will be to develop autonomous robotic systems that can perform complex tasks in human environments and safely cooperate with humans in arbitrary settings. Robots with these capabilities will transform our everyday lives, as well as industrial processes, like the Internet, cell phones and computers had in the past two decades. Many globally-relevant agencies and institutes, such as the International Federation of Robotics (IFR), McKinsey Global Institute (MGI), and World Economic Forum (WEF), have highlighted the significance of the field of autonomous, next-generation robotics, and its potential to transform life, business, and the global economy. The IFR 2018 executive summary reported a robot sales increase of 30%, which is a new peak for the fifth year in a row. WEF has placed Social robotics as one of the top emerging technologies of 2019. The MGI report on disruptive technologies estimates the potential economic impact on an annual basis by 2025. For Advanced robotics, it could range between $1.7 -$4.5 trillion, while for Autonomous and near-autonomous vehicles estimates range from $200 billion to $1.9 trillion. Robotics has been a high-technology area for decades and competitive advantages therein are hard-won. European robotics research strategies clearly indicate that Europe must not only retain leadership where competitive advantages have been achieved, but also take the lead in first-wave technologies gaining valuable intellectual property rights and first-to-market advantages.

Artificial Intelligence (AI) has made tremendous progress since the chess-playing system of the 90s that could defeat leading human experts. Even though contemporary and most publicly covered results are still in the domain of board games, AI algorithms are tirelessly working hard in the background giving us more relevant search results and suggestion, analyzing complex and large amounts of data, reading radiology scans and assisting diagnosis, or simply making sure that we take that perfect photograph. Just like robotics, AI will be one of the main technologies underpinning technological, economic and societal growth, and just like robotics, it is likely to be the competitive advantage of the 21st century. It is forecasted that AI could greatly contribute to the global economy, not just from increased productivity, but also due to the increased consumption of the new AI-derived products. WEF top emerging technologies from 2017 to 2019 always include AI, e.g., as part of solving visual tasks, robotics or producing AI-led design. The MGI report states that Automation of knowledge work, thanks to advances in AI, could have a potential economic impact on an annual basis by 2025 between $5.2 -$6.7 trillion. Given that, AI presents an enormous opportunity and a grand challenge, since in such competitive domains first-to-market advantages present big pay-offs and are hard-won. Therefore, as in robotics, Europe should not afford to be complacent and must take the lead in first-wave AI technologies to ensure valuable advantages.

Although the initial dream of a fully autonomous robot that can operate in an unknown, unmodified environment is still present and driving robotics research, the users in factories, hospitals, and homes are still waiting for the autonomous robots that researchers promised long ago to deliver. A smart autonomous robot that can adapt to changing conditions and operate in dynamic human environments needs a sophisticated level of intelligence to function correctly. Such an interplay of perception, control and reasoning of the physical agent and intelligent algorithms, yields a paradigmatic challenge that needs to be addressed – the convergence of AI and robotics. This convergence will produce new operational capabilities: industrial robots that can work safely alongside humans without fences, social robots that can assist the elderly and physically impaired, autonomous vehicles that can spare us of routine drives, and finally allow us to stop sending humans to do the robot’s job. In essence, robotics research needs advanced AI solutions that can make future robots generalize to unpredictable situations and reason in a timely manner; but also, autonomous machines that will be capable to interact and work safely with humans give rise to specific, robotics-oriented challenges for the contemporary AI. All this gives rise to and motivates the field of study of Artificial Intelligence for Robotics. However, these intelligent robots need to produce results that are ethical, socially responsible and human-centered.

The recent document published by the European Commission on Trustworthy AI states that Trustworthy AI has three components that should be met throughout the system’s entire life cycle; namely, that it should be lawful, ethical and robust. In AIFORS, we focus on the technical aspects of achieving Trustworthy AI in the field of robotics, which dominantly applies to the robust component:

“[...] individuals and society must also be confident that AI systems will not cause any unintentional harm. Such systems should perform in a safe, secure and reliable manner [...], important to ensure that AI systems are robust [...] both from a technical perspective (ensuring the system’s technical robustness as appropriate in a given context, such as the application domain or life cycle phase), and from a social perspective (in due consideration of the context and environment in which the system operates).”

Furthermore, the document states, among the technical methods for realizing Trustworthy AI, that for a system to be trustworthy, we must be able to understand why it behaved a certain way and why it provided a given interpretation. This is particularly important for pervasive robotic systems working in symbiosis with people and their environments using AI algorithms for decision making. Thus, in order to better understand the system’s underlying mechanism, we need Explainable AI (XAI), i.e., AI that will be robust, efficient and interpretable, while maintaining a high level of performance. The involved methods are vital not only to explain the robot’s behavior to users but also to deploy reliable robotics technology, i.e. transparent and accountable autonomous robotic systems.