Find out how to Grow to be Future-Prepared By way of R&D Experimentation


Any R&D undertaking is an iterative course of the place we now have preliminary hypotheses and begin constructing software program elements to validate them or have classes realized for knowledgeable decision-making. R&D tasks are cyclic as we should revisit assumptions a number of instances, pivot, enhance, right – it sometimes takes a couple of iterations earlier than getting the true consequence.

Vik Bogdanov Hacker Noon profile picture

Vik Bogdanov

Tech storyteller

Final 12 months, I wrote an article for HackerNoon about distant R&D and the way it may help companies speed up their tempo to innovation. This 12 months, I’ve revisited the subject and are available throughout some mind-blowing stats and info concerning the world R&D market, which impressed me to discover additional:  

  • What sort of corporations are higher positioned to take a position and profit from R&D,  
  • What kinds of tasks and options are greatest fitted to R&D, and  
  • How R&D may help drive Trade 4.0 transformations, increase the evolution of rising applied sciences, and allow extra enterprises and SMBs to develop into tech-savvy, modern, and future-ready. 

To reply these questions and to higher perceive how R&D tasks differ from mainstream software program engineering tasks, I’ve talked to Bogdan Popescu, Technical Director of R&D Embedded Division at, a Romanian know-how options firm. 


Bogdan has greater than ten years of hands-on expertise with R&D undertaking improvement and has made his means from R&D Program Supervisor to R&D PM to Head of R&D. Particularly, he focuses on managing R&D tasks throughout AI/ML, Virtualization, and IoT/embedded domains. 

Bogdan, might you please describe your present function as R&D Technical Director? What’re your key priorities? 

My present function covers two foremost areas. On the one hand, I’m in command of all tasks operating in our R&D and answerable for their supply to prospects, assembly deadlines and high quality standards, constructing belief with prospects, and paving the way in which for long-term collaboration. That’s my foremost focus.   

Nevertheless, alternatively, my job is to establish potential new know-how use circumstances which may be capable of assist different corporations clear up their tech limitation points and have a constructive impression on their enterprise progress. From this angle, we construct proof-of-concept (PoC) tasks, which indicate a excessive stage of experimentation and enjoying round with applied sciences and instruments. 

What’s the present dimension of your R&D workforce, and the place do you supply software program engineering expertise for it? 

At the moment, we now have round 60 individuals in our R&D workforce which are concerned in present buyer tasks and PoC tasks improvement. For buyer tasks, we sometimes recruit engineers from our out there expertise swimming pools in Jap Europe.

For PoC tasks, we don’t rent everlasting assets. We decide individuals with required abilities from present buyer groups (who briefly have a restricted work scope or need to volunteer to work on R&D tasks as their pet tasks) to run PoC experiments. They see it as a solution to advance their data and increase skilled progress, as most PoC tasks are very difficult and require non-trivial approaches. 

What’s the important thing distinction between mainstream software program improvement tasks and R&D tasks? 

I’d say that every R&D undertaking is kind of a terra incognita for us. Mainstream tasks are usually well-defined, whereas in R&D tasks, we now have many unknowns, we do not know of the ultimate undertaking scope and dimension. We might (or might not) know the vacation spot, however we now have no clue find out how to get there.  

As such, any R&D undertaking is an iterative course of the place we now have preliminary hypotheses and begin constructing software program elements to validate them or have classes realized for knowledgeable decision-making. R&D tasks are cyclic as we should revisit assumptions a number of instances, pivot, enhance, right – it sometimes takes a couple of iterations earlier than getting the true consequence.  

That is an instance of a digital twin PoC resolution constructed at R&D that makes use of HoloLens to allow distant management of the robotic arm.

How is the client finances allotted for R&D tasks? 

In our R&D, we use a lean, step-by-step strategy, i.e., we begin small and scale. As such, the finances is added incrementally because the scope turns into clearer.  

What’s your key takeaway from R&D experiments? 

Our PoC tasks assist construct long-term strategic relationships with our prospects. After they come to us with nice enterprise concepts (that couldn’t be dropped at life beforehand on account of technological limitations) and see us discover a workaround experimentally – they’re amazed. Because of this, they get extra confidence in our capabilities. 

One other takeaway is – you don’t must underestimate R&D tasks. I’ve seen many instances in my profession when a undertaking that begins as a small PoC, one ultimately transforms into a strong software program engineering undertaking on account of new concepts and options being recognized. 

Do you’re feeling that your company R&D division is competing with R&Ds of enormous tech corporations for locating and constructing principally the identical tech options? 

I wouldn’t say there is direct competitors between us. As we don’t know what options different R&Ds are engaged on, it’s onerous to inform if we’re making an attempt to construct the identical issues or not.  

Our strategy is relevant to industrial domains largely. We construct and ship business industrial options. We don’t experiment for the sake of experiments. In the beginning, we experiment so as to add worth to our shopper engagements. Tech giants can afford to launch R&Ds simply to complement tutorial analysis on the most recent applied sciences or to construct new merchandise from scratch, spending enormous budgets. 

We do have plans to contribute to tutorial data by sharing our PoCs and their use circumstances. By the way in which, a while in the past, our R&D workforce contributed to Intel’s white paper on deep studying. 

So, whereas we now have a unique R&D focus, the philosophy is analogous: we attempt to construct new issues with present know-how. 

R&D helps companies flip their present tech limitations into future enterprise alternatives, capitalize on innovation, and achieve a powerful aggressive benefit.

On the customer aspect, what sort of corporations outsource their R&D tasks essentially the most? 

Largely enterprises. They outsource their innovation tasks to keep away from heavy upfront investments in staffing and retaining an costly in-house R&D workforce, and constructing inner experience and area data. Enterprise tech decision-makers select to piggyback on third occasion experience, as they admire getting exterior suggestions earlier than asking for an even bigger finances and committing to full-fledged product improvement.  


One other group of corporations that leverages our R&D capabilities is SMBs. Their buyer journey with us begins once they see our R&D PoC demos and notice that’s precisely the answer that may assist repair their most vital drawback or differentiate available in the market. They need us to conduct extra in-depth analysis, and that’s once they resolve to outsource their pilot R&D undertaking.  

Startups, nevertheless, hardly ever use exterior R&D. For many of them, a know-how product is on the coronary heart of their enterprise and monetization mannequin, so they like to build up data and develop deep area experience in-house.  

How do you deal with shopper’s mental property (IP) in R&D tasks? 

We’ve got blended approaches. Every little thing is determined by the undertaking. For some tasks, we grant full IP rights to the shopper. For others, we now have shared IP. Let’s say the shopper involves us with ready-made software program elements, and we solely do the mixing work. On this case, software program elements IP will belong to them, and we’ll hold IP for the mixing resolution we’ve constructed. 

Wrapping up, the evolution of software program improvement instruments and processes and the rise of 5G, IoT, edge computing, and quantum computing will additional push integrations, deployments, and seek for new enterprise fashions, applied sciences and requirements – and that’s the place R&D could be the primary catalyst of change.  

Vik Bogdanov Hacker Noon profile picture


Be part of Hacker Midday