Orlando Avila GarcíaSoftware Architect
Relation to current technological trends:
Autonomous Systems – In order to provide higher levels of autonomy to systems performing in increasingly dynamic uncertain complex environments, such as autonomous cars, basic sensory-motor loops need to be enhanced with advanced and adaptive perception, decision and action processes. We are interested in the kind of intelligent behavior that emerges from the interaction between the system’s components and its environment. This is a biologically-inspired and bottom-up approach to AI and intelligent robotics:
- Behavior-based robotics
- Intelligent perception and action
- Reactive planning
- Decisional autonomy
Cognitive Robotics – Cognitive neuroscience has progressively provided insights about the kind of physiological mechanisms underlying affective and cognitive phenomena in humans and animals. The simulation of these processes has become a key element to achieve autonomy in robots and systems to ensure that they remain viable and safe. This inspiration in the actual working of the brain lets us depart from more engineered and symbolic (top-down) approaches to AI, such as knowledge-based and expert systems:
- Human-robot interaction
- Affective social robotics
- Interpretation and imagination
- Cognitive robot architectures
Development and Epigenetic Robotics – A modern AI-based system acquires knowledge from its experience in a continuous interaction with its environment. Beyond pre-defined knowledge bases and rules engineered by humans, those systems can learn the best representations and course of action to better attain their goals. Self-organization and collective intelligence, as well as cooperation with humans are more complex but still effective mechanisms for adaptation to dynamic uncertain complex environments:
- Deep reinforcement learning
- Collective intelligence / robotics
- Collaborative robots (cobots)
Robotic Systems Development – Successful real-world autonomous robot solutions face important challenges in their design, construction, testing, deployment, operation and evolution. This is related to the increasing level of distribution and complexity of the so-called cyber-physical systems. Traditional software and systems engineering concepts, methods and tools need to be combined with AI development ones. Among our topics of interest we find:
- Software and systems architecture
- Systems design and integration
- MDE of complex systems
- Collective adaptive systems
Robot and Robotic System Safety - Increasing levels of AI in “smart” sensory-motor loops allow intelligent robots and systems to make decisions with increasing degrees of autonomy, with human beings progressively ruled out from the control loop. AI safety is a rapidly evolving field that has become one of the key components of a successful deployment of intelligent autonomous robots. We work actively in the following topics:
- Robot ethics
- Safety by design
- Risk assessment and reduction
- Human-robot interaction safety
- In ARI, we have experts in the design, construction and deployment of automated decision making processes in different application domains, including self-adaptive and self-managed application and services.
- The IAR&S Research Line activity is enhanced by a strong collaboration with the NG Cloud lab to develop the concepts of cloud robotics and edge robotics which aim to offload robotic tasks from the physical robot and deploy them along the compute continuum (edge to cloud).
- We collaborate in the H2020 CLASS project (2018-2021) which seeks to develop a novel software architecture to support connected car applications, such as intelligent traffic management and advanced driving assistance https://class-project.eu/.
- The market is interested in advanced research on affective computing mechanisms to embed safety guarantees in intelligent autonomous robots and systems for operation in proximity to and/or in collaboration with humans.
- Collaboration in the creation of the AI safety engineering ecosystem, bringing together traditional safety engineering and AI safety communities. Co-organization of the First International Workshop on AI Safety Engineering (WAISE2018) https://www.waise2018.com/
- Artificial Intelligence (AI) technologies enable a digital future where products and services deliver people-first experiences to customers. AI adds an adaptive layer on top of existing IT and addresses the complexity arising from the interplay between Big Data, IoT and Cyber-Physical Systems (such as autonomous cars).
- IAR&S Research Line contributes to shaping the future Atos Industry 4.0 and IoT offering. It envisions intelligent autonomous systems and robots that streamline real-world operations in a wide range of industries by means of real-time automated data-driven decisions and actions.