AI IN DYNAMIC ACTION
about as...
The AI in dynamic action (AIDA) scientific group is dedicated to the study and application of artificial intelligence in dynamical environments, with a special focus on safety and trustworthiness. Among various machine learning approaches, we are especially interested in reinforcement learning - a methodology resembling the action of living beings in changing, uncertain environments which react by punishment and reward. Applied to human economy, AI has to fulfill requirements on safety and trustworthiness, especially regarding the personal data privacy. These requirements become particularly challenging in dynamical application of AI, such as robotic, autonomous driving, medical therapy support, chemical engineering, energy management etc.
The scientific group seeks to apply an interdisciplinary approach by the fusion of machine learning with various fields, such as system and control theory, to develop novel dynamic AI methods.
The team of the scientific group
Assistant Professor
Head of scientific group
Head of scientific group

Georgiy Malania
Junior Research Scientist

Ilya Osokin
Junior Research Scientist

Vadim Artyomov
MSc students

Vyacheslav Kovalev
MSc students

Mikhail Patrikeev
Engineers

Kirill Myasoyedov
Engineers
Projects & research


Computer vision system for Advanced Driver Assistance Systems (ADAS).
We develop and test algorithms based on computer vision for road lane recognition and driver warning, prevention of leaving a virtual corridor is integrated as well.
Rcognita
Rcognita is a flexibly configurable framework for agent-enviroment simulation with a menu of predictive and safe reinforcement learning controllers. A detailed documentation is available here.github.com/AIDynamicAction/rcognita

Control algorithms for stable locomotion of quadrupedal robots.
We are develop algorithms for stable locomotion of leg robots based on model predictive control and reinforcement learning with special stabilizing machinery. This will allow stable walking especially walking up and down stairs of legged robotic platforms for better usability in civil, and Industrial environments.
Design, assembling implementation of a Greenhouse mobile robot for tomato monitoring.
We develop and deploy a Greenhouse robot with computer systems and data processing systems on board. This systems enable the robot to navigate between rows and capable of monitoring tomato condition.Publications
Beckenbach, P. Osinenko and S. Streif. A Q-learning predictive control scheme with guaranteed stability.European Journal of Control 56 (2020): 167-178 |
NO INFO |
Beckenbach, P. Osinenko and S. Streif. On closed-loop stability of model predictive controllers with learning costs.European Control Conference, 2020 |
NO INFO |
Osinenko, A. Kobelski and S. Streif. A method of online traction parameter identification and mapping.IFAC World Congress, 2020 |
NO INFO |
Osinenko, L. Beckenbach, T. Göhrt and S. Streif. A reinforcement learning method with real-time closed-loop stability guarantee.IFAC World Congress, 2020 |
NO INFO |
Russwurm, P. Osinenko and S. Streif. Optimal control of centrifugal spreader.IFAC World Congress, 2020 |
NO INFO |