Skip to content Search

The role of drones in today’s world

Unmanned aerial vehicles (UAVs), also commonly known as drones, are becoming increasingly common in various aspects of life. Over the past few years, drones have become more accessible and technologically advanced, making it possible to conduct research combining artificial intelligence and aerial observation.

One of the most common areas in which drones are used is industry. For example, they can be used to inspect buildings, bridges, chimneys, or critical energy infrastructure, enabling damage requiring repair to be identified of or any deviations from nominal parameters to be monitored. Drones are also used for inspections of photovoltaic installations or wind turbines, enabling their efficiency to be monitored and faults in hard-to-reach places to be detected, where regular inspections are costly and time-consuming. In medicine, drones are used to deliver medicine and other medical supplies to places that are difficult or impossible to reach by any other means. They can also be used for rescue purposes such as searching for missing people in the mountains or at sea, even in difficult weather conditions, thereby increasing safety and the range of available sources of information.

However, before drones become a permanent part of our sky, a number of problems related to safety and the absence of a human pilot must be solved. At IDEAS NCBR, we are conducting research on autonomous operations, in which the drone receives only general instructions or commands such as “find the injured person,” and all direct flight decisions must be made independently based on data obtained from various sensors.

Autonomous operations

To fully utilize the potential of drones, we need to solve all of the key problems that prevent us from deploying them even in the most unpredictable conditions and ensure the reliability and efficiency of control systems. The most important aspect of flight operations is safety as drones must operate independently of the environment, weather conditions, and situations they encounter. Another challenge is optimizing the battery level, which is essential for their autonomous operation. We must ensure that drones perform tasks efficiently while maintaining sufficient energy reserves to complete the mission by a safe margin. The optimization of missions enables maximum resource utilization and minimizes the risk of mission interruption due to battery depletion.

At IDEAS NCBR, we focus on creating advanced computer systems and artificial intelligence algorithms, using the latest technologies in machine learning and computer vision. Our innovative approach enables the creation of drone management systems that will effectively coordinate their operations, even in the most demanding situations. One of our priorities is to classify the current situation, both nominal and emergency, for drone behavior to be as well adapted to specific conditions as possible. This way, we will achieve autonomous, yet safe operation for both the environment and the drones themselves.

Safe transfer from the laboratory to the sky

Before drones can perform operations near people and buildings, rather than under prepared laboratory conditions, a number of mechanical, electronic, and programming problems must be solved. One way to ensure safety is to duplicate systems, (known as redundancy), so that in the event of a system failure, a backup system provides the desired functionality. As far as software reliability is concerned, almost all aspects of modern unmanned aerial vehicles can be developed and tested in simulated environments. Unfortunately, for some problems, the solution that seems effective in a simulated environment turns out to be difficult or even impossible to implement in the real world. These situations are called sim-to-real problems and are one of the biggest challenges for robotics and artificial intelligence specialists. Although simulations can precisely plan and test their operation, in the real world previously unanticipated factors may arise, some of which cannot be reliably simulated. Therefore, to ensure the reliability and effectiveness of our solutions, they must be gradually phased into use and continuously tested under real conditions, not just theoretical ones.

Research Team Leader


Karol Pieniący

Karol Pieniący is a PhD student at the University of Warsaw’s Faculty of Mathematics, Information Technology and Mechanics. He has been developing his experience in robotics and automation since 2010. He is an expert in unmanned aerial systems, dedicating much of his doctoral work to designing, building and programming drones and ground elements of autonomous systems. His specialized knowledge and experience in the field enable him to leverage the latest technologies, resulting in both efficient and innovative systems.

He has worked hands-on and as a consultant on all types of robots, including wheeled, walking, swimming and underwater robots, production lines, manipulators, fixed-wings, and multirotors with electric, combustion, and hybrid drives. He specializes in system architecture (especially distributed systems), system integration, project management, and team management.

Currently, as a PhD student at the University of Warsaw, he is conducting research on a multimodal project combining a drone and a robotic manipulator. Additionally, he also runs a course on Robot Control for Master’s students at the University of Warsaw. On this course, he imparts his knowledge and experience to students, helping them understand issues related to designing and implementing control algorithms for different types of robots, effectively encouraging them to tackle modern robotics problems.

Karol loves competition. Even before starting his studies, he began competing in robotics competitions, including autonomous robot contests organized by the Kaunas University of Technology. During his studies, he continued his passion by participating in Mars rover competitions and drone races, in which he also achieved many victories.

In 2019, he founded the Warsaw MIMotaurs team, which he still leads as captain. The team has had successes in various competitions, but their greatest distinction was second place in the Machine Learning category in the prestigious AlphaPilot competition. As part of this event, teams designed autonomous drones capable of flying through a specially prepared course. Their drone, equipped with innovative machine learning algorithms, achieved the third-best time in the final race.

The team’s latest success is entering the finals of the international MBZIRC competition, which is organized by ASPIRE from the United Arab Emirates. The team passed the qualifications and simulation phase and is now among the top five teams from around the world, who will compete for victory in February 2024. Not only does the MBZIRC final give the victor a prestigious $2,000,000 prize; it is also an opportunity to gain knowledge and exchange experiences with other teams from around the world.

Other research groups and teams

  • Algorithms in Autonomous UAVs Unmanned aerial vehicles (UAVs), also commonly known as drones, are becoming increasingly common in various aspects of life.
    Karol Pieniący
  • Sustainable Computer Vision For Autonomous Machines Our solutions could potentially be used in drones as a tool supporting the protection of national parks, including animals against poaching. They allow for fast and efficient monitoring of large land areas in remote locations...
    Bartosz Zieliński
  • Learning in Control, Graphs and Networks The team develops neural networks that generate graphs. These solutions are oriented towards the automatic design of structures
    Paweł Wawrzyński