İlk Kez Uzayda Yapı İnşa Eden Robot: 3 Seviyede İngilizce Haber

A robot built part of a structure in space. No astronauts helped it. According to Caltech, this was a first. Caltech is a US university. It shared the news on 5 November 2026.

The project is called ASTRA. A US defence agency called DARPA paid for it. ASTRA went to space on a rocket on 30 March 2026. It flew into low orbit around the Earth.

The robot's arm picks up tubes and joints and puts them together. The plan was to build an antenna 1.4 meters wide. The full antenna needs 11 sections. ASTRA went to space with two sections. In space, the robot built a third section.

The mission ends in December. The team will probably not finish the whole antenna. But the test showed that the idea works.

The engineers talked to the robot about seven or eight times a day. Each time was about 15 minutes. They had a twin robot in their lab. When there was a problem, they tested a fix on the twin first.

Very big things do not fit in a rocket. In the future, robots could build very big things in space. These could be telescopes, antennas and solar panels.

For the first time, a robot has assembled part of a structure in space without the help of astronauts. The robot was designed and built by researchers at Caltech, the California Institute of Technology, which announced the result on 5 November 2026.

The project, called ASTRA, was funded by DARPA, a US defence research agency. It was launched on 30 March 2026 on a SpaceX rocket and went into low-Earth orbit, together with nine other payloads on the same vehicle.

The payload is roughly one cubic meter in size. Inside, a robotic arm picks up tubes and joints from motorized dispensers and puts them together into repeating sections called bays. These bays form a frame known as a truss. The final goal was a reflector antenna 1.4 meters wide, which needs 11 bays. ASTRA left Earth with two bays already built, and in orbit the robot added a third.

The work was not easy. Engineers on the ground could contact the robot only seven or eight times a day, for about 15 minutes each time, and they checked photos from space before sending new commands. When something went wrong, they recreated the problem on an identical twin robot in their lab, tested a fix there and then sent it up.

Because of time limits, the full antenna will probably not be finished before the mission ends in December. However, the team says the experiment showed that the method works.

This matters because rockets limit the size of what we can send to space. Building in orbit could one day make huge telescopes, large antennas and very large solar panels possible.

Until now, anyone hoping to put a large structure into space has faced a stubborn constraint: it had to fit inside a rocket. Engineers have worked around this in two ways, either by folding a structure for launch and unfolding it in orbit, or by relying on astronauts to piece it together by hand. A small robot built at Caltech has now demonstrated a third option.

According to Caltech, which announced the result on 5 November 2026, its robotic truss builder became the first machine to assemble a structure in space autonomously. The project, known as ASTRA (Autonomous In-Space Tensioned Reflector Assembly), was funded by DARPA through its NOM4D program and lifted off on 30 March 2026 from Vandenberg Space Force Base, riding to low-Earth orbit as one of ten payloads aboard Momentus's Vigoride 7 vehicle.

Packed into roughly a cubic meter, the payload contains a robotic arm that draws composite tubes and aluminum joints from motorized dispensers and fits them into repeating sections, or bays, of a supporting truss. The ultimate aim was a 1.4-meter reflector antenna requiring 11 bays. ASTRA launched with two already in place, and the robot completed a third in orbit.

Operating it was hardly straightforward. As Caltech describes it, the team could reach the robot only seven or eight times a day, in windows of about 15 minutes, and studied images from orbit before issuing each command. To troubleshoot faults, they reproduced them on an identical twin in the lab, tested a solution there and only then transmitted it to space.

Time limitations meant that only the first new section could be built, and with the mission ending in December, the full truss will probably remain unfinished. Nevertheless, the researchers regard the experiment as a proof of concept.

The ambitions behind it are far larger. According to New Atlas, DARPA program manager Andrew Detor said the program aims for structures hundreds of meters or even kilometers across, adding that nothing remotely close to that scale can be built today. If the approach scales up, it could pave the way for vast telescope mirrors, large communication antennas and enormous solar arrays gathering power in space.