Interview

A model of reality

Dr Linas Gelazanskas, Director, UAB IT logika (DRONETEAM), talks to Danielle Kenneally 

Creating a 3D city-scale reality model for the city of Vilnius

WELL-TRAVELLED and well-educated, Dr Linas Gelazanskas has honed his knowledge of electronic and computer systems engineering to provide the city of his birth, Vilnius in Lithuania, infrastructure monitoring. Studying at Lancaster University in the UK with a year spent at North Carolina State University in the USA, Dr Gelazanskas returned to Lancaster to complete a PhD in engineering on smart grids and AI, all while embarking on a career using drones.

This resulted in the development of a DBOX system, a drone-in-a-box, that can create 3D city-scale photogrammetry. Its endless applications for the city of Vilnius provides unrivalled support for the municipality’s needs and in some cases is a lifeline when used by emergency responders.

If that was not enough, Dr Gelazanskas is involved in many more projects besides, including drone entertainment, LightaSky, as well as being a Vilnius Gediminas Technical University professor-partner. It is fair to say, he is adept at achieving anything he puts his mind to.

Here, he talks to the Civil Engineering Surveyor’s Danielle Kenneally about his largest and most challenging project so far with the municipality of Vilnius.

Tell us about UAB IT logika (DRONETEAM) and the Vilnius 3D city-scale photogrammetry and reality modelling project which was also your Bentley Systems 2023 Year In Infrastructure and Going Digital Awards entry.

DRONETEAM.lt is a dedicated team of drone professionals. We have more than 10 years’ experience of providing photogrammetry on both small scale objects, like art, to large scale environments. This is one of the main services we provide with our DBOX system, which is a drone-in-a-box system developed by us, as well as LightaSky, which is the use of light drones as an entertainment performance.

For our Bentley Systems 2023 Year In Infrastructure and Going Digital Awards entry we submitted one of our DBOX system projects based in Lithuania. The project uses drones flown by city-based pilots alongside a team who process the captured data using Bentley software, such as OpenCities Planner and iTwin Capture, to create a 3D city-scale reality model for the city of Vilnius. The ongoing project collects data autonomously and remotely using AI detection to process continuous changes into a 3D model of the city.

What is the aim of the project?

There are a lot of applications for this project. This includes for municipality owned companies, such as refuse collectors, construction sites, and the emergency services, can monitor and be monitored using this technology.
Another important application of the system is for emergency first responders where drones can be flown to the location and make an assessment of the scene. This can range from helping firefighters to locate where to begin, to tracking burglars. 

For example, not only can fly-tippers be caught using AI and 3D modelling to locate the item previously seen elsewhere in the city, refuse collectors can use the system for waste management optimisation. They themselves are given three days to remove the fly-tipped item or items by the municipality or they can be fined. The system can also be used to monitor dilapidated, at risk buildings. This is especially important in relation to dangerous cracks and unstable roofs that could collapse at any time. Plus, it can be used to monitor construction sites, as well as observe the work is in keeping with not only the submitted plans, but the expected quality and timeframe too.

Another important application of the system is for emergency first responders whereby drones can be flown to the location of the emergency and make an ongoing assessment of the scene. This can range from helping firefighters to locate where to begin, to tracking burglars. There are so many examples. To put it simply, this project is all about city infrastructure monitoring.

How is Bentley software being used within the project?

Bentley Systems software enables us to create 3D models and develop insights using AI detection. The photogrammetry we capture using the drones is processed using MicroStation and iTwin Capture. These construct the city-scale reality models which can be found and continuously updated in OpenCities Planner to develop insights for the city. Furthermore, we use LumenRT to demonstrate these models and their value to the community to the public.

Did you have to overcome any challenges within this project?

As with many cities, there are flying regulations in Vilnius with which we had to follow. It took us nine months to complete our first flight permit. To do this we had to coordinate with the Lithuanian Transport Safety Administration, as well as the Lithuanian Civil Aviation Administration (CAA). It was an monumentous task but extremely worth it.

I am also currently part of the air navigation service provider team where we are in the process of developing a user-based system to allow drone pilots access to a quicker knowledge based hub. Drone pilots will be able to ask a question in relation to their flights and have an answer within 20 minutes rather than the current nine months. This actually took us nine months to teach, as well as learn in Lithuania, in accordance to the European Authority for Aviation Safety (EASA) regulation (EU) 2019/947 on the rules and procedures for the operation of unmanned aircraft. Once in place it will enable drones and planes to coexist in the same airspace.

What benefits does 3D city-scale photogrammetry and reality modelling bring to stakeholders and in particular this project?

3D city-scale photogrammetry and reality modelling provides so many benefits for its stakeholders. At its heart, and in particular for this project, it is not only about city infrastructure monitoring but also city management optimisation. Each element of this project has its own benefit. So for example, construction site monitoring has been estimated to have saved € 30m during the early phases, such as through error detection. Then there’s also waste management optimisation, which benefits from cost savings and performance improvements. Plus, there’s also an element of saving lives when working in tandem with emergency services and the police, as well as crime prevention. The benefits are vast.

How is digital advancement helping cities address other societal needs, such as agriculture and security?

To put it simply, this project is all about city infrastructure monitoring.For us at IT logika, we have special multispectral cameras which support crop management. We can produce maps of normalised differential vegetation index (NDVI) for agricultural scientists. These can help make decisions on crop spraying ideals, such as what kind of substrates to use – peat moss, liquid manure, etcetera – alongside other factors like the time of year. I’m not a specialist but there is an ecosystem being born right now. Farmers buy this data to better understand their crops and what they need.

In terms of security, some of the biggest security firms in the Baltics are using our systems. They have informed us that the digital advancement of drones is helping them to save on in-person patrolling. This could be a construction site, a factory, anywhere motion sensor drones to monitor the area. All in all, the possibilities for digital advancements are limitless.

Have advances in reality modelling driven the growth of city digital twins?

Definitely – when I attend conferences, like the Bentley Systems 2023 Year In Infrastructure and Going Digital Awards, I can see that reality modelling is being recognised more and more. Everyone is now using not only lidar for scanning which is reality modelling, but also photogrammetry. It is advancing at an incredible pace, its speed and ease of use means that more and more companies and individuals are turning to these techniques to support their work. It is not simply for city digital twins but for any objects and areas, large or small. In fact, this conference is a great platform to showcase this, as a presentation given by Greg Bentley showed that 44% of the finalists were utilising iTwin to create digital twins of a number of environments.

What will we see next in 3D city-scale photogrammetry and reality modelling?

Speed and ease of use will increase, of course, and AI will continue to progress. I also think we’ll see a more robust classification of objects being developed, both from 2D and then when it is developed into 3D. This will enable us to clean the model more easily and refine it. We’ll also be able to seamlessly rescan parts of the city to be able to blend it into previous scans to create a 4D model. It’s exciting to see where it will go next and also how we will develop with it. 

Dr Linas Gelazanskas, Director, UAB IT logika (DRONETEAM), talks to Danielle Kenneally

linas@droneteam.lt

www.droneteam.lt