Two Portuguese students invented a robot that replants forests destroyed by fire

Two 19-year-old Portuguese innovators have developed a new all-terrain reforestation robot designed to plant trees in forest fire-ravaged areas that are too steep or dangerous for people or heavy machinery to reach.

The forest was the intimate setting of his childhood. Marta Bernardino and Sebastião Mendonçatwo 19 year olds, who while growing up near Lisbon, always believed that forest would be a constant in their lives.

“It was a living amusement parkwhere we built worlds, a sanctuary where concepts of ‘importance’ were felt instinctively, rather than being taught”, says Marta.

But every year, the two young men saw the fires ravage the forests close to their homes, leaving behind gray and burnt slopes, says .

Desperate to revitalize these forests, the two, then high school students, decided to create — a six-legged robot capable of reaching and reforesting areas where humans were unable to reach.

“Trees are vital: They provide us with oxygen, store carbon, stabilize the soil and provide a home for the world’s wildlife. For the health of our planet and its inhabitants, its preservation and protection must be a priority for all of us”, highlights Marta in an article on .

“The main reason why reforestation is not being achieved, besides poor forest management, is that human work brings many challenges in terms of injuries, maintenance and planting capacity”, explains the young student.

Portugal is one of the European countries most affected by forest fires. One published in 2024 by atmospheric scientist Carlos C. DaCamarafrom the University of Lisbon, concluded that more than 1.2 million acres burned between 1980 and 2023, the equivalent of 54% percent of the territory.

Only in 2017, the country lost 32,000 acres of tree coverwith fires responsible for 75% of this destruction. Many of these fires occur in areas with rugged terrain, so steep that neither volunteers nor teams of forestry sappers can access them.

“The steep terrain prevents manual planting and the arrival of machines heavy burden on most of the burned areas in Portugal”, the two young people recently explained in a presentation of the project.

With more than 60% of Portuguese forests Located on difficult slopes, conventional reforestation methods have had difficulty keeping up with the recurrence of fires and soil degradation.

Students began prototyping the Troubadour in 2023. The first prototype, which it cost 15 euros and was built with recycled partsplanted small trees 28 percent faster than humans, with a reported survival rate of 90 percent and no need for post-planting care.

This first result led the duo to develop a more robust versioncapable of operating autonomously on slopes of up to 45 degrees, Marta tells Smithsonian Magazine. “We build all-terrain robots that carry small trees on their backs and plant them independently on difficult terrain.”

The Troubadour’s hexapod design Distribute weight evenly on single hairavoiding compaction associated with tractors or other heavy vehicles, which can hinder water infiltration and the supply of oxygen to the roots.

The robot uses a depth camera to map obstacles and adjust your trajectory in real time, while an AI system analyzes pH and soil moisture before starting a three-step planting sequence: excavate, place, compact.

The team claims that this method has already been validated, achieving survival rates of 85 to 90% in field trials.

Have been tested drones for aerial seedingbut these disperse thousands of seeds per acre with low precision. “Drones spread seeds with little precision — wasting one of the scarcest natural resources“, notes Marta. Some pilot projects recorded survival rates between 0 and 20%.

The Troubadour, instead, plant one by one small trees that are already rootedchoosing micro-niches where the probability of survival is greater, which reduces waste.

With capacity for plant up to 200 young trees per hourTrovador also sends GPS coordinates, soil data and battery information to the cloud, allowing remote monitoring. Future updates should allow the robot automatically avoid areas that are too dry and focus on areas with the best growth potential.

The cost of building this small robot is one of the concerns of the two young inventors — who, however, plan to respond to this challenge by offering the Troubadour as a service and not as a hardware product.

In the model they propose, municipalities, insurance companies, forestry companies and NGOs can design an area in an application, choose native species and receive a budget for “trees on the ground”.

The team expects this approach to be significantly cheaper than manual planting and more cost-efficient than drone methods, if seed wastage is taken into account.

The goal of the two young people is to have Trovador operate in large forest regeneration projects by 2026, creating, in Marta’s words, a reforestation “fast, accurate, auditable and scalable to millions of hectares that climate models say we need to restore this decade.”