Five Steps to Launch an Own Robotic Project

In 2018, Rozum Robotics turns two years old. By now, the company has built a collaborative robot and found its first customers. In this article, the Rozum Robotics team shares the five steps to launch a successful robotic project and common mistakes to avoid.


The PULSE collaborative robotic arm, or simply cobot

The team says, the biggest secret is to be really into robotics—otherwise, all the five steps are useless. You need to believe, follow your dream, and to be eager to change the world. The rest is nothing without it.

At the start, you will have to face numerous challenges and problems. Without inspiration, the risk is too high that you’ll leave it halfway.

The market

Before you start implementing a project, you should find out what the market really needs. According to statistics, 42% of startups fail only because no one needs their products.

The robotics market is huge, comprising industrial robots, drones, as well as service and household robots. (By the way, sex robots are a hot trend, today.)

Rozum Robotics set eyes on the emerging market of collaborative robots (cobots) as a prospective niche. Unlike big and heavy industrial robots, cobots can work in close proximity to a human and are absolutely safe.

A demonstration of a collaborative robot

The company conducted extensive research and analyzed a massive array of data, including demand levels, potential consumers, purchase and service costs, competitors, etc.

The chosen niche seemed to be promising, and so it is—analysts predict the market to grow a hundredfold within ten years. Cobots are times cheaper than industrial robots. New algorithms and improved design of software interfaces have made programming easy, even for a non-professional.

Two years ago, competitors were not numerous. So, Rozum Robotics had no doubts about the choice, determined to build a cobot of its own.

The people

People make all the difference. Finding people with required skills is a task of foremost importance.

In contrast to Germany and the USA, the robotization level in Belarus is low. So, searching HeadHunter or similar websites will hardly get you a tenth fraction of the staff you need.

Rozum Robotics tried a lot to solve the problem. Firstly, the team attended specialized conferences and competitions in robotics. There, one can meet young specialists—their eyes burning and heads full of ideas and knowledge.

Most probably, you will have to teach them, though. You will teach them and learn yourselves—on a continuous basis. The company’s specialists still have to swallow three–four books a month, just to keep up their skills. Rozum Robotics even has a specialized library. If you’re interested, you can request a subscription.

Another idea for talent acquisition is to visit higher education establishments. In Belarus, these are the Belarusian State University (the Department of Theoretical and Applied Mechanics), the Belarusian University of Informatics and Radioelectronics (the Department of Control Systems), and the Belarusian National Technical University (the Department of Robots and Robotic Systems).

Every year, these establishments graduate over a hundred specialists in robotics. Unfortunately, only ten out of the hundred graduates are motivated and skilled enough to engage in robotics professionally.

Rozum Robotics is still hunting for talents. Every week, candidates come to the company’s office for job interviews. You are also welcome, if you have genuine interest in robotics. Perhaps, you can find a place where you really belong.

The scope of works

Before you start designing a product, you need to define clearly what components you need and where you can get them. Essentially, there are three options:

  • to do everything on your own
  • to buy everything for big money
  • to buy everything cheap


For buying cheap, your destination is AliExpress. At first, you will be excited about the variety and the prices, but disappointment is right around the corner. Real-life tests show that the motors are no good: the components start to come off in just five minutes of operation.

Then, you google for more, to find top-grade manufacturers. You are excited again! You get in touch with them, and they redirect you to a dealer. The dealer tells you the price. The price upsets you. But you have to try at least, right? You order—just to get even more upset.

Your order is too small, which means a higher price and three to six months of waiting for delivery. Then it turns out that the part you need has a non-standard size. Customization costs are so expensive that you wonder, maybe, you misheard the amount or miscounted the zeros.

Rozum Robotics has tried all the three options, finally settling with building its own frameless and servo motors.

Rozum Robotics presented the products in Europe and the USA. Despite initially intended as components of a robot, the frameless motors and servos are now sold as separate products. The company started selling them even before it commenced the sales of cobots.

The team says, if they’d found some hypothetic “rozum robotics” company that could deliver servos for their robot at an acceptable price within reasonable terms, there would have been no need to invent something.

Now, the industry got one more player. For its part, Rozum Robotics is ready to support new startups and share their technologies.


Remember: the earlier you start talking to potential customers, the better. Show the MVP to as many people as possible. Any feedback is valuable. It may prompt you a problem you missed or, vice versa, alleviate your concerns.

For those who want to commercialize a new product, the Rozum Robotics team recommends reading the “The Four Steps to the Epiphany” by Steve Blank. The book outlines the Lean Customer Development theory.

You can understand what the market really needs only through communicating with consumers. There are no ready-made solutions, though. No one can tell you straight what they’d like. People will tell you things, of course, but not the things you need to know.

Most of us are simply used to the same routine in everyday life—in household or at work. It’s like this: people know that a car needs a gear box. So, if you ask a driver, he or she’d tell you that they’d like a more dynamic gear box with more or fewer speeds, an automatic gear, a variator, or the sport mode. Hardly anyone will tell you they need a car without a gear box. We’re all thinking within a paradigm.

So, you should ask potential consumers to describe the conventional process: what and why a person does, etc. This is the only way to see all the nuances. The technique is one of the many offered under the approach.

After Rozum Robotics finished building a robot, the team set off for international exhibitions and specialized conferences.

In 2017, the company attended major exhibitions in Russia, Germany, the Netherlands, and the USA. The experience was useful, first of all, in terms of marketing and sales. At the exhibitions, Rozum Robotics met its first customers.

The company also got a chance to talk to end users. Letting them see and interact with the robot, the team got valuable feedback. Now, the company is about to release a new version of the robot, improved based on the comments from consumers.

The sooner you present your product, the better, even if it has only minimum functionality. You should keep it in mind that the product is never final. It’s constantly evolving and improving. However, you should start showing the product when you are still a little uncomfortable about it.


The Sales

Another task of foremost importance is to hire a sales manager at the initial stage, right after an idea of a project comes to your mind. Ideally, you should have at least a hundred fully paid up pre-orders by the time you start the design stage.

If you have a cool product (otherwise, why bother doing it at all?), you should be able to convey its benefits to your customers, so that they would be grateful for being the first ones to buy it.

You’ll have to choose a specific sales model. You can either earn money through sales, or servicing, or options.

In robotics, the conventional model is as follows: the manufacturer—the dealer that delivers a robot to another country—the integrator who installs the robot at a manufacturing site—the end user.


The conventional sales model

Rozum Robotics wants to simplify the chain so as to reduce the price. Eliminating the integrator will make adoption of a robot easier, to the extent it is possible. You should also remember: the more local dealers you have, the better coverage you will get, and the faster your product will get to end users.

In the end, Rozum Robotics came up with the following model:

     Option        Advantage


Firstly, one can buy a robot via the company’s website.




This breaks the existing pattern. You register with the website, get access to information, and pay with a card or from an account. The delivery service will get the robot right to your door.


Secondly, the company continues working with dealers.


  Rozum Robotics offered their vendors a significant discount and informed about the recommended prices, warning that if the price will differ, consumers might prefer to buy via the website.
  Thirdly, the integration.  

Rozum Robotics is working hard to prepare a lot of text and video instructions telling and showing how to assemble, commission, set up, and program a robot. The software is also simple. Consumers can also ask for remote support by a developer or call an authorized dealer-integrator (an additional source of income for dealers).


The current model is neither the best nor the ideal one. It is just different—one of a dozen possible options. Time will show whether it will be a success.


The sales model tested by Rozum Robotics

So, those were the five steps to build a commercially viable robot. Perhaps, you’ll argue it takes the same five steps to create any commercially viable product—and you are absolutely right.

Robotics is not only and not so much about technology, but rather about business. To do the things you love doing, changing the world for the better, you need to make your business successful. As for the technical details—about the torque and the moment of inertia—you can easily learn that from books.

So, go ahead and build robots. Believe in yourself, your team. Try and fail, then try again. Next time you can sure do better.

And remember: no matter what your current result is, next time make it better or just in a different way

Original version of the article


Have any question or challenge we can help with? Feel free to contact our team!

Contact Rozum Robotics

1 /*modal window*/ /**/