Cheap drones capable of autonomous control can now threaten billions of euros worth of technology and are fundamentally changing the nature of modern conflicts. The need to develop, test and produce drones and systems against them more quickly is now a strong voice in European security debates. “We have talented students and scientists, but the Czech Republic does not have sufficient financial support,” says Martin Saska, head of the Multi-Robot Systems Group at the Czech Technical University.
We hear about drones and autonomous systems very often, but rarely specifically. What are the most fundamental innovations that are changing modern warfare today? The main thing that has changed is that much smaller drones, multi-rotor drones, have also begun to be deployed. It is no longer just the large winged aircraft that flew long distances and monitored. Now these small drones are also capable of searching for and killing individual soldiers, which is a big change.
In Ukraine, people describe how they watch FPV drones in the sky every day. Are the huge numbers, swarms and the ability to kill individual soldiers part of the main transformation of drones? We are not talking about real swarms yet. They are a large number of individually controlled drones that are controlled by a human. Swarm technology could do much more. That number is precisely the change that has occurred, thanks to which the battlefield has completely changed.
There is a lot of talk about drones controlled by artificial intelligence. How far is this technology really today? According to my information, this technology has been deployed on the battlefield at a minimum. Scientific circles, some university teams and private companies can do much more with artificial intelligence.
We are actually able to build a flying robot that flies completely alone, reacts to the environment and has all the computing power on board. Such sophisticated drones have not yet been deployed in Ukraine. First of all, because the technology they use is not that far advanced, but also because these systems are more expensive.
European leaders are calling for the acceleration of drone production and implementation. But Ukraine is showing that new systems are being created and modified within weeks. Can Europe even do enough to transition to practice? The most important thing is to create a critical level of knowledge in the state, to put together sufficiently large teams of experts that cover all aspects of autonomy. This critical mass of people is then able to respond very quickly and flexibly to the needs of change.
It is not about immediately throwing oneself at one specific problem that can be solved differently tomorrow, but about building sufficiently large companies here that will have up-to-date know-how for solving problems and will be able to dynamically develop and improve their products on a much larger scale than is the case now.
Lack of resources
Connecting experts in one place with manufacturing companies would probably be the best, but how long does implementation take today, where are the biggest obstacles? It is about resources. I had the opportunity to visit China, or rather the professors I work with, they receive huge funds directly from the government, from Beijing, to build a spin-off of a university team and thus be able to create products much faster.
I'm not saying that we don't have support, there is support for applied research, startups and companies, but it is much smaller. For example, co-participation is always required, so I have to get money from the state in parallel, which is significantly more demanding than I could see in China. At the same time, I still have to find money to finance all of it.
It's all complicated, there are relatively long delays, waiting for a grant, for an increase in money, this is where we are really running out of steam. I'm sounding the alarm a lot now, looking for private financing, private investors, because if we don't manage to build something really big in our region, a big team that will be able to make really competitive products, then in the current world in this technology we have no chance.
When there is money, will there be anyone to produce such developed new systems? Are there sufficient capacities here to do it very quickly? If we are talking about the Czech Republic, for some reason there are a lot of skilled students and scientists here, the human potential here is enormous, but also the production potential. I have now set myself the goal of developing a drone within the company we are working with that will use completely Czech components, or those from Central Europe. And we have succeeded.
We have our own smart batteries, sensors. The only thing we cannot do is chips and computers, but Europe is working on that too, so I would not be afraid of it, the potential here is great.
This is also related to the fact that we have managed to build one of the largest research groups at CTU that do drone research, perhaps even in the world. We are the largest in Europe and we can beat teams from all over the world and Europe in direct competitions.
What are you working on specifically now? What could be used for defense or the military? Right now we are focusing on those completely autonomous swarms that are controlled by on-board computers, that can operate in the field without communication, without GPS. In that case, I can practically not disturb them in any way, I don't need to pull those optical fibers there, which is a technology that is decades old, which with is now using in Ukraine.
We can do it without that and we are also focusing on deploying artificial intelligence on board so that the control of drone teams is not, as it is now, via a joystick, at all. The robot must perceive its environment, but the commands must also be in natural language. It is not only about defense, but also about helping in many other applications, for example, in searching for lost persons, in helping the police, firefighters.
There are many applications in the civilian sector and everywhere there is a need for users to communicate with the robotic system naturally. Police officers are not robots and they are not supposed to be. But they should have the ability to say: “A child got lost somewhere in the forest, a person was seen walking in this direction,” and the robotic team will evaluate this information and based on that react or propose a solution completely on its own.