Alumni Interview # July 2020

What was the title of your master thesis project and who was your supervisor?

I did my thesis work under the guidance of Dr. T. Keviczky.

The title was: Roll stability control of autonomous truck combinations

What are your main responsibilities as a Robotics Software Engineer at Mainblades?

It is mainly twofold. On one hand, I work on several parts of the onboard software – being a control geek, it’s mostly related to the autonomous flight of our quad-copters. On the other hand, I recently picked up a project aiming to solve the problem called full coverage path planning, applied to our use case, which is aircraft inspections.

Working at a startup means that you are not confined to a narrow set of topics. Sometimes you do hardware-related stuff or “get your hands dirty” by going out to the field to test – at Mainblades, it is mostly airline hangars.

What does a regular workweek look like for you?

Our teams work in scrum, which gives a nice skeleton for our days and weeks. Each day we start with team stand-up meetings, exchanging information about our progress, issues, ideas that might affect the whole team. After that we usually do reviews for each other, if needed – this means understanding what your teammate implemented and testing/approving it before it could become production code. Subsequently, you proceed with the issues of the current sprint, according to the priorities the whole team agreed upon during the common sprint plannings. Nevertheless, as the company is heavily focused on R&D, most of our tasks involve lots of uncertainties that require frequent refinements in our planning. These refinement meetings take place weekly.

What do you like most about your work?

Being really close to the fire. Everything you do here directly impacts the business. Having worked more than two years at large multinationals, I truly recognize and appreciate the difference – it is vast. It also means that you have a larger saying in the company’s matters. Furthermore, it is lovely to work in a company where everybody is sort of my age. Seeing my algorithms in action provides an extra gratification.

What do you like least about your work?

Getting in flow while creating a functioning product is definitely fun – taken to an extreme however, it will have a negative impact on one’s work/life balance. This again can affect productivity. I constantly have to remind myself of this, because the amount of work is literally never-ending.

Could you tell us about an interesting or recent project you did?

Our product is a technology that – among other things – enables a drone do navigate around an aircraft autonomously. To do this, we continuously scan the aircraft’s surface with LiDAR and utilize a simultaneous localization and mapping (SLAM) algorithm to determine the drone’s position in a GPS-denied environment. SLAM, on the other hand, will never have such a high update rate as GPS and our MPC controller does not like a reference trajectory that is jumping around a lot every now and then. Marrying the two components demanded a lot of creativity. An additional challenge compared to university projects is the safety-critical application. One expensive mistake at a customer could have a devastating consequence on the whole company.

How did you become interested in this field?

Autonomous systems already attracted me when I chose my thesis topic. This field offers a very nice balance between the needed theory and implementation practices. Besides, I like to see my work in action.

What advice would you give someone a student who is considering this type of job?

Just go for it. Your 20’s are the best times to be in a start-up environment!

 What steps would you recommend to take to get into this field?

System & Control does, in my opinion, a pretty good job at educating you in a certain aspect. However, if you wish to work on a product that has to sell, do not forget about educating yourself towards software. Learn what it takes to write production-level code and you will stand out.