DEUS: Deep Understanding of the Situation
- Project title: Deep Understanding of the Situation
- Project acronym: DEUS
- Funding source: FNR - Luxembourg National Research Fund
- Funding scheme: FNR-CORE-2022 - Junior track
- Project id: C22/IS/17387634
- Project starting date: 01 Sep. 2023
- Project duration: 36 months
- Funded amount: 495.000 Euros
- Project role: PI
Abstract:
Robotics plays a pivotal role in the strategy to expand the competitiveness of all sectors of the economy, as well as offering new solutions to societal challenges. While industrial robots are typically performing repetitive tasks, operating in industrial settings where the world is designed around them, service robots assist humans at work in non-industrial settings or in the home, needing to operate in complex dynamic unstructured environments. Robots need to continuously acquire a complete situational awareness in such environments (i.e. perceiving the environment within time and space, comprehending its meaning, and projecting it in the future) to enable intelligent decision-making and autonomous tasks execution.In this project, named DEUS (which means God in Latin), we focus on the development of a novel robotic situational awareness (ROSA) as it is an essential missing CORE capability on autonomous and intelligent robots which is blocking their further development. Our novel ROSA will pursue the goal of deeply understanding the situation with multiple levels of abstraction such as geometric (e.g. shape of the objects), semantic (e.g. type of the objects), or relational (e.g. relationships between the objects), integrating the perceived measurements from multiple sensor sources, with past observations and background knowledge, and it will align with the requirements of accuracy (i.e. properly representing the real situation), efficiency (e.g. computationally tractable and real-time capable), robustness and versatility (e.g. handling novel situations or applications), dependability and trustworthiness (i.e. available, reliable, and safe), and explainability (i.e. comprehensible by a human).
This project is rooted in the latest state of the art on machine learning -based and optimization-based situational understanding for robots. While the former methods are essential to excerpt deep knowledge, they are usually not capable of performing in real-time or building an explainable, versatile, and trustworthy long-term understanding of the situation as the latter, which normally provide a shallow situational understanding.
We will develop a ROSA that combines the best of the two worlds. Machine learning -based algorithms will be adapted to comply with the needs of robotics systems, and optimization-based algorithms will be extended to incorporate deeper situational understanding. Our research will focus on two different aspects. On the first hand, we will research how to acquire a deep long-term situational understanding from the sensor measurements. Concretely, we will concentrate on both semantic-geometric and semantic-relational knowledge. On the other hand, we will research how to acquire deep long-term semantic-relational situational understanding by reasoning on the knowledge that the robot already has.
Our research will not only be experimentally validated using our ready-to-use robotic platforms, but to prove its technological relevance, we will demonstrate a proof-of-concept in two real-world use-cases we have selected, namely a UAS for surveillance of critical infrastructure, and a walking robot for inspection of construction sites.
Besides the scientific impact that a PhD thesis, our open access publications, and our datasets will generate, guided by our committee of stakeholders, we expect that the DEUS project will bring high long-term socio-economical impact, by providing the foundations of an essential component of the robotic artificial intelligence that will enable a limitless number of potential applications for service robots.
Related Research: