Where do you think cobotisation in facility management has moved in the last two years?

In terms of technology, the most important thing is that robot navigation systems have improved significantly. Today, we have the first machines that share information, thus communicate with each other and continuously update a map of the environment and exchange it. This increases both the accuracy and efficiency of their operation. In addition, cleaning robots today clean much better than before, leading to a higher quality of cleaning.

How can I imagine such communication between robots?

Previously, we had to map each machine separately and then manually transfer and edit the maps. Today, we map one machine and the others just „connect“. They all share the same version and continuously update it together. For example, if one of them repeatedly hits an obstacle, it automatically writes it into the map and the others can then avoid it.

What about the robot market itself – do you see any changes?

Yes, there are a lot of new manufacturers, but it’s often similar to the automotive industry when big brands close partnership with Chinese or Taiwanese manufacturers. In terms of quality of service and spare parts, it is still more profitable to work with European partners. Another area is the subject of safety, which two years ago was a rather marginal issue, but today it is a key topic. Robot safety is essential for us. We are considering what is safe, what is no longer safe, and how to set up a system that cannot be remotely hacked or misused. Of course, it also depends on the type of operation. In some places there is no sensitive data, while in others there is a need for thorough security.

So who guarantees the security of the robots, the manufacturer or the operator?

Partly the manufacturers, but very often we customize the robots and their systems according to the client’s requirements. The cloud in which the robot operates also plays a crucial role. Today we have several levels of security and the client can choose:

• The basic variant – the robot operates in a native environment with common security parameters.
• Private cloud – data flows through a secure internal environment, which significantly increases protection.
• Offline operation – for some clients we choose a mode where the robot is completely disconnected after startup and operates fully offline.

Are clients already thinking more about automation or is it still very much about education?

A significant number of clients already see the benefits of automation – as do many of our competitors. They realize that finding reliable employees for cleaning positions is becoming more challenging. When the war in Ukraine is over, the situation may get worse due to the outflow of labour. It is therefore wise to be prepared. At OKIN Facility, we are now focusing on dry cleaning, which is more sustainable and efficient.

What exactly does „dry cleaning“ mean?

Simply said, it is a form of vacuuming. We currently use two methods. The first is similar to conventional vacuuming – first the surface is brushed and then the loose dirt is vacuumed away. A less conventional, but just the right method for industrial premises, are robots that use centrifugal force instead of suction, called „airflow“. Larger dirt bounces off the brush and smaller dirt is carried by the airflow. These particles stop at the fan, which is protected by a HEPA filter. Here they stop and fall into the waste bin, which is often much larger than a normal bag. This system is ideal for dusty and industrial environments.

Do you find that clients prefer a „wet clean“?

Yes, we sometimes encounter the idea that it’s not „clean enough“ without water. But sustainability-conscious clients directly request dry cleaning, as water and chemicals are two major issues. Washing machines with 150-200 litre tanks use several tanks in a single night, so the final consumption is huge. Moreover, it is not always necessary. But again, it’s about assessing the suitability of the premises. Dry cleaning works best in warehouses with pallets or cardboard packaging. There, dry cleaning is more efficient, cheaper and more economical. Moisture is not desirable in these areas as water could damage goods or disrupt the environment which is deliberately dehumidified. Additionally, in extremely dusty environments, wet cleaning can create near-mud that clogs machines.

Do you also use dry cleaning in offices?

Yes. There we often deploy small robots that have rotary brushes. Unlike conventional vacuums that only take a surface, these machines can deep comb the ballast carpet, wring out hair and capture fine dust with a HEPA filter. This too contributes to a better and healthier environment – like our robot at headquarters. Thanks to our experience and knowledge, we always choose the right technology for the type of space. There is no one-size-fits-all solution, and that’s where our advantage lies. We can combine technologies to suit each client’s needs.

In an earlier article, we talked about design barriers for autonomous robots (cables lying on the floor, backpacks placed next to a table). Are buildings and interiors starting to adapt to the needs of machines?

From my perspective, not much has changed yet, but there has been significantly more discussion around „robot-ready“ buildings. When companies build a new headquarters, they are now more open to discussions about, for example, having a cleaning room where the robot can be cleaned. This was often forgotten before. Unfortunately, offices in general are still not very robot-friendly, dealing with chairs, opening doors, etc. Technically this is solvable, for example a robot can open a door, but it requires the installation of an electric opener, which is more expensive than just programming the function. Moreover, the program can be easily replicated, while the opener needs to be physically fitted at each door.

So it will still be a while before buildings are „robot-ready“?

Some architects are actually starting to think about it. We know of several projects that are being planned along these lines. But the majority of buildings? That’s a question for the next 50 to 70 years, maybe more. So either we adapt to robots, or we teach robots to cope better with existing environments. That is why we are moving towards our vision of humanoids, which is taking on more shape because it is now possible to buy a humanoid.

How would you specifically use a humanoid in facility services?

I personally have believed in humanoids for years. That’s also why I’ve long had a picture in my presentation where we start with a bucket and a mop and end with a humanoid taking the bucket and mop. And it’s becoming relevant. You can buy a humanoid these days. One Chinese company has launched a 160 cm tall humanoid model that is more powerful and can be programmed. Although it can’t do much in its basic form (it can walk, run, clap), its technical design is very impressive.

However, the biggest benefit I see in simple, repeatable tasks. For example, walking into a meeting room, taking things off the table, dusting, pushing back chairs, and opening the door for a robot to clean the floor. That’s a realistic goal. It may take 6-12 months to teach a humanoid to do such a task, but once it’s done, other robots will pick up the knowledge. And suddenly there could be 20 of them, all doing the same thing as the first one.

Is it difficult to teach him such a task?

The fundamental weakness is object recognition. A humanoid can’t tell the difference between a plastic and a glass bottle. And that’s a problem – because he doesn’t know how to grip it firmly. It can’t even tell if a glass is full or empty. In addition, the camera’s vision is still poor at detecting transparent objects.

When I look at the picture of the humanoid, I’m surprised to see that its face is disembodied. Is there a reason for that?

It’s intentional, because people are generally afraid of humanoids that look human.

I would also like to mention that we also cooperate with Czech Technical University, where they are working on the topic of humanoid-caregiver. However, I personally tend to the idea that the humanoid belongs in the service, to do an activity that is routine or physically demanding for humans, and in turn, this could result in more people going into care services.

Could this one day backfire on us?

Of course – just as a car can be dangerous in the wrong hands. It’s about being responsible. But once we’ve mastered humanoids, the next question is: what’s next? Maybe clones of humans who clean up perfectly. It really does feel like a science fiction movie – but development is moving forward all the time.

Where do you see robotics going in the next two to three years?

Definitely towards humanoids. I trust them – just because of the increasing computing power and better image recognition capabilities. I think we can adapt them in a few years for common service tasks such as cleaning, carrying objects or preparing a space. Their advantage over conventional single-purpose robots is precisely their multi-purpose capability. When the single-purpose robot has finished its job, it remains standing. But the humanoid can move on to another task. Its an investition in one „brain“, one logic, but we can use it to its full potential in different situations.

What advice would you give to companies that are considering robotic cleaning but are still hesitant? Maybe because they don’t have enough information?

Meet us.