Innovative Solutions to Existing Problems

Case Study: Reach-U

We sat down with Gaspar Anton, the Head of the Mobile Mapping Division, in Reach-U to find out what they’ve been doing, how they’ve been doing it and how they’ve ended up with an incredibly innovative method for applying UAV’s in the field of mobile mapping and we of course wanted to know why & how they ended up with the e-Drone Zero.

  1. For those less familiar with Reach-U please provide a short description about your company

Reach-U combines the brightest minds of geoinformatics, data scientists and software engineers under the unified single goal – develop what nobody has, do what others can’t and make the innovation happen where most of them fail. One of our world leading product lines is DEMOGRAFT – a mobile locational intelligence based big data analytics platform for Mobile operators, smart cities, etc.

Living by the same single goal my adventure has been like no other. My name is Gaspar Anton and I’m the founder and manager of “EyeVi” mapping technology platform. Now being a department of Reach-U, it’s incorrect calling myself a founder, but this describes my situation the best. For 3 years I have been leading the team of enthusiastic geniuses to build up something that will shape the future. In all this Reach-U has provided us with a helping hand as an incubator, start-up hub and the first investor simultaneously while letting us grow at our own pace.

  1. We are always hearing the word “Smart City” these days, where does your mobile mapping platform “EyeVi” fit into the typical Smart City workflow.

Well, EyeVi is what every city needs to boost up the growth and move towards becoming the future Smart Cities. The need for accurate and detailed geodata has become essential in various segments. Utility and communication networks, infrastructure, buildings, vegetation, advertisements, billboards, traffic signs, sidewalks, street lights, parking slots and so on – basically everything that exists in the city environment will soon be digitalized if not already. And this is all because of the private or public companies, associations or other entities operating in the city environment that have started to understand the true benefits geoanalysis. If 5 years ago it was enough to know the shape of the building in a map application then now the details of visual, measurements, connections, land usage, surroundings and any other smallest details up to actual 3D shape is what everybody is looking for.

  1. So you’re going to be building 3D cities? That must be a massive amount of data to collect.

The main challenges regarding this is not only how to acquire all this content but even more on how to keep everything up to date. Most of the geodata is being currently managed by outdoor teams who use some equipment or just mobile apps to go around and collect the needed information. This however is both very time-consuming and expensive when factoring in yearly costs to keep the info up to date. The other option used is the aerial scanning, but it comes with other issues such as limited final accuracy levels as well as heavy costs for constant updates.

We are providing an alternative – creating a high accuracy 3D environment as a base and using neural network algorithms to extract valuable detailed information for any segment. The base data is collected and updated with a car mounted mapping system consisting of a panoramic camera, LIDAR scanner, GNSS/IMU system and other sensors. The full solution includes data collection, AI based automated features extraction and end user applications for data management – or as we call it, “EyeVi” Mobile Mapping based visual intelligence platform.

  1. Who is your usual target segment in those areas and how do you expanding your reach while staying ahead of the game?

Our current client base consists mainly of 3 target segments, starting from the local governments or state road authorities to whom we are providing the input of road attributes, objects and pavement defects for effective road asset management, secondly the infrastructure engineers with the need of high accuracy classified 3D information for whom we are providing the input dataflow of road profiles, DEM (digital elevation model), DSM (digital surface model) or other specifics through API-s and as for the utility, energy or telecommunications network planning and design our automatization can give enough detailed input to calculate accurate construction costs and provide options for optimization. And these 3 are still only a section of the actual geodata market segment with a lot of other ongoing activities. Still the limits of the current solutions used are preventing the actual growth of this segment. The key here is to offer everybody what they need and at the same time sharing the costs of acquiring it all.

Our current plans are doing exactly this – building up our visual intelligence platform in every major city in Europe and providing its benefits to different targets through API-s or end-user applications. Our mapping fleet will keep the base data constantly up to date and with every new use case we will improve our information extraction selections and capabilities.

  1. What are the current issues that you face which could hold you back the most in the future perspective?

Operating in the mobile mapping field for more than 8 years we know its strengths and limits quite well – based on that we have also selected our current 3 main target segments. However, the city environment is where the car-based mapping limitations are best noticed, especially by the users whose expectations are not met. This includes everyone who would like to look beyond the city streets – buildings, parks, yards, pedestrian areas and so on. Or everyone who needs to observe a specific area, city block or even entire city as a whole. The mobile mapping setup can provide maximum information from the street level, but not the full 3D details of the city – and we do know how much it is needed by the different users of the same target groups we already cover.

  1. We wouldn’t be talking if you hadn’t looked into UAV’s being a solution to these issues, but please clarify how you arrived at this decision and how it all fits together.

Yes, to fulfil this gap we are looking towards drones. I mean who wouldn’t. In the last 2-3 years the commercial UAV segment has had a massive growth and in the upcoming years it doesn’t seem to slow down, rather speed up even more. Already now the mapping equipment available for drones has become more accurate and with constantly decreasing prices. By using UAV-s it would be possible to scan the entire city fast and effectively having the maximum coverage and detail of the 3D visual environment. However, most of the mapping grade drones that are in the expected cost range have several preventive technical aspects that limit the usage of UAV-s for us. Maintaining the required accuracy level is one of these. With the airborne LIDAR systems, the angular error from IMU (Inertial Measurement Unit) from long distances and oscillating winds have a much bigger effect. For compensation much more accurate equipment should be used compared to Mobile Mapping Systems. But by upgrading to this level you will raise the total unit cost 3-4 times if not even more making it once again an unreasonable solutions when calculating the operational costs for constantly updatable dataflow production.

We are planning to work around this issue with a combination of different inputs – generating the maximum accuracy level 3D base data with car mounted technology and to ensure the quality it will be adjusted with every available dataset of geodetic survey or measured control points. Then we use the same automated matching algorithms to add the full input from drones. The result being an engineering grade 3D base data set with maximised city environment coverage. Now imagine building the visual intelligence platform on top of this input!

  1. What are the key requirements for you in a drone suitable for your vision?

I have been keeping my eye on the drone market for quite a while now because not every option fits in with our ambitious next step plans. The target we have set for ourselves differs a lot from most of the usual projects. Our success rate will be very dependent on our fast and effective scaling. This means that we are not planning to not only cover a specific area or the full city, but also to maintain it all and implement our platform with constant data generation processes in every city in Europe. Within our 2 year execution plan we produce the base data for our platform covering up to 1 million miles of road network and to be successful with these massive amounts we must maximize our efficiency throughout our entire workflow, especially in the base data generation part.

We have already done it with our mobile mapping technology and now are looking to do the same with the drone-based data collection system. There is 1 main aspect that I am always calculating when evaluating any UAV systems and that’s their operational timeframe. With the expensive equipment set the total covered areas per one unit in month, week or day has a major effect on the overall cost structure especially while targeting a big territory. In terms of drones the main parameters affecting the operational timeframe are of course flight times and limiting the unavoidable down-time not to mention the reliability of that technology as well as its capabilities for autonomous tasks execution and safety to everyone around. As a quick example: raising the operational timeframe by 1h a day, from the usual 5h to 6h, you could use 4 times more expensive equipment and still have a lower overall cost in comparison. And this already applies when targeting only 10 cities so consider what effect it has with 100 or 1000.

  1. So why the e-Drone Zero? It does have perhaps the more riskier / less proven technology in it after all.

I must say I was very impressed and excited when I first heard the plans of SKYCORP’s CEO, Marek Alliksoo, to build the next generation drone that will outperform other solutions in a similar level by a very good margin. It almost sounded too good to be true, at least that was what everybody was telling me. From one innovator to another I knew that everything is possible with the right ingredients and strong determination. At the same time, I was really hoping it was possible because of the massive effect it would have to my expansion plans. And I’m glad I was waiting for that since they did it and expanded the flight time by more than 3 times using the next-gen hydrogen fuel cell tech. Looking at the drone you can see that every core element was carefully selected to provide the maximum efficiency in all areas. And what I love the most is that the full UAV platform is modular giving the opportunity to add and replace additional components. In addition the highly advanced brain inside it, a well-developed AI,  gives the unit top-quality senses in terms of visual recognition, movement, multitasking and of course safety.

  1. Tell us more about your plans now or how they might change accordingly.

With several projects already running and many to come, we are currently re-evaluating every Mobile Mapping use case. We have set one question to guide us in this phase of R&D, “What if we do it with e-Drone Zero”. So yes, we are testing out the full capacity of this unit to find out where is the optimal balance between MMS and UAV based mapping solutions and how it will change in years to come. I know that the e-Drone Zero has a bright future ahead and I really hope that the regulations for the use of UAV systems will be built up as well so it would be possible to run the system and whatever will follow to their full capabilities.

Thank you very much for taking the time to talk with us Gaspar and provide such detailed answers. We wish you all the best with your plans and of course hope to see what you can do with the e-Drone Zero!