Market signals

The construction site started generating data

Gustavo Santos·25 September 2026·4 min read

A video circulating on social media says "your bricklayer turned into a Chinese robot". It carries 420 kilograms, works for hours and never takes a day off.

The phrase is good for going viral. But it hides the most interesting part of the news.

What actually arrived

The Chinese company BrightMaster Robotics (BMR), from Shenzhen, has been manufacturing robots for civil construction since 2018. According to the company itself, it has already delivered more than 5,200 units, used in more than 1,600 construction sites and in more than 40 million square meters built, across 13 countries.

In Brazil, the exclusive representation belongs to Grupo Minas Brisa, a mining group from Minas Gerais, which also takes the line to Argentina and Mexico. The official launch was at Concrete Show 2026, in São Paulo, in late August.

The line presented includes robots for:

  • sanding of plaster, with dust extraction;
  • interior and facade painting;
  • laser leveling of fresh concrete;
  • logistics: autonomous transport of up to 420 kg of material;
  • measurement: 3D scanning of the construction site within a radius of up to 400 meters;
  • cleaning and intelligent construction elevator.

The promises, and whose they are

According to Grupo Minas Brisa, each robot is equivalent to 3 or 4 workers, can be up to 3 times faster than manual execution and should reduce labor costs by around 30%. The equipment costs from R$ 250,000 to R$ 1 million, with an estimated return after around 10,000 m² completed.

All those numbers are from the supplier. That does not make them false. But they are not yet evidence for your construction site. And the viral video that talks about "hundreds of square meters per hour" provides no source; we did not find that number in the company's information.

Context helps to understand the interest: entities such as CBIC and FGV IBRE have been pointing to a shortage of skilled labor in construction, with project delays and rising costs.

The robot is not the news

In Data-Driven Culture, construction is a classic example of a physical operation: the signal is out in the world (deadlines, waste, rework, downtime) and almost never turns into data. What the site foreman sees stays in his head, in a notebook or on WhatsApp.

Look at the list of robots. One of them does not paint or carry anything. It measures. It scans the construction site in 3D.

And all the others, in order to function, record what they do: how many square meters, in how much time, with how much material, where.

That is what changes. For the first time, part of the construction site begins to produce records automatically. What was a dispersed signal begins to turn into data.

What can be done with this

With the project on one side and the scanning on the other, the construction site gains a comparison that previously depended on a tape measure and goodwill: what was planned and what was executed.

This has a name in theory: management by deviation. The plan is a hypothesis. The actual is reality's answer. The gap between the two is a signal. Of an error, of a wrong assumption, or just of normal variation.

Where this can go wrong

Technology scales the culture that already exists.

In a construction company that discusses deviation every week, with a responsible party and a decision, the robots' data can turn into intelligence. In a construction company where reports serve to assign blame, they will turn into yet another dashboard that nobody opens. Visibility is not consequence.

There is another risk, more silent. If only the supplier's technician knows how to operate, calibrate and interpret the equipment, the construction site trades the dependency on the good foreman for dependency on the technician. It is the same captive intelligence, with a different uniform.

How to know if it is right for your construction site

The question this site keeps asking applies here too: how do you know?

Before believing in the 30% or the 3x, it is possible to test:

  • measure the baseline first: m² per day per team, rework, paint and plaster consumption, accidents and sick leave;
  • run a pilot on similar fronts, one with robots and one without;
  • add the total cost: operation, maintenance, energy, logistics, downtime and learning curve, not just the price of the machine;
  • look at quality, not just speed: flatness, thickness, touch-ups.

And remember the proportional dose. If the return comes after around 10,000 m², a small renovation may not need a robot. It needs, first of all, to record what happens.

What remains

The news is not that the bricklayer turned into a robot.

It is that the construction site, one of the places where reality speaks loudest and is least recorded, started generating data.

The question now is the same as always: what will construction companies do with them?