The Budapest‑founded company greehill now helps cities from Singapore to São Paulo assess and manage their urban trees using LiDAR surveys and proprietary artificial intelligence. However, the technology is not yet applied in Budapest: the founders say the barrier is budgetary prioritization rather than a lack of local expertise.
What greehill does
The company’s mission is to make urban tree stocks healthier and safer and to maximize the trees’ ecosystem services — shade, urban cooling, carbon sequestration, particle filtration and improved quality of life. greehill uses LiDAR scanners and AI to establish the current condition of trees, identify which trees require intervention, and repeat surveys every two to three years to monitor trends.
LiDAR produces high‑resolution point clouds from which tree attributes can be automatically extracted — species, size, vitality and trunk stability under different wind loads.
Origins: Singapore and R&D
The company’s early work dates back to 2014, when high‑resolution LiDAR surveys were conducted in Budapest. That expertise led them to Singapore, where NParks became a partner and initiated a local consortium R&D project involving the space agency, SZTAKI, the land office and universities. The R&D succeeded and turned into multi‑year contracts: the initial five‑year engagement created a digital platform that underpins Singapore’s tree management; a new ten‑year cycle followed.
Preserve existing trees as well as plant new ones
The founders stress that a tree needs 15–20 years to reach maturity, so planting alone does not solve near‑term problems. They cite that currently about 50 percent of trees live in climate zones they are not adapted to, and that figure could reach 75 percent by 2030. Without timely investment, greehill warns, many cities may have little left to save within five years due to climate impacts.
Concrete projects and incidents
- São Paulo: after a major storm, 1.2 million people lost power for four days because many trees fell onto power lines. greehill secured the assignment within four months and has produced a detailed tree cadastre to support management.
- Mumbai: three months earlier a falling tree hit a school bus and an 11‑year‑old child died — an example cited of reactive policy after tragedy.
- Rome, Singapore and others: hundreds of thousands of trees have been scanned in pilot and operational programs; in Singapore some areas are surveyed annually, others biennially.
Technical and organizational challenges in scaling up
greehill is active in about 130 cities across five continents. A major scaling challenge is data volume: surveying thousands of kilometers of urban networks generates large daily data flows — the founders report roughly 4–5 terabytes per day — which must be transmitted, processed and reprocessed. Another critical challenge is automation: machine learning models must achieve high accuracy to evaluate trees automatically and economically.
Human and organizational challenges remain as well: a relatively small team must manage global logistics and client relationships. Working with the public sector implies long sales cycles — from initial meeting to first revenue commonly takes 1.5–2 years on average.
Which AI do they use?
greehill has developed its own discriminative AI since 2017 rather than relying on generative models. Their system recognizes and characterizes trees: species recognition reaches about 90 percent accuracy, while feeding the same data to a generative AI produced 23–30 percent accuracy. The founders emphasize that model errors can have safety implications — for example misjudging trunk stability under wind could lead to serious incidents — so they built a bespoke, robust technology stack.
Internally, the company is transitioning its entire operation to be AI‑driven not to replace staff but to boost throughput: AI 'agents' will handle routine tasks while specialists focus on decisions and complex analyses. Their target is a 30–100 percent productivity increase with the current headcount.
Funding, investors and business model
The company reports more than EUR 25 million invested in R&D. Its investors are wealthy private individuals rather than institutional funds; these backers accept longer time horizons appropriate for selling to the public sector, where procurement is slow but yields stable long‑term contracts.
A key element of the business model is automation: the more processes can be reliably automated, the more scalable and cost‑effective the service becomes.
Why keep the headquarters in Budapest if the city does not use the technology?
The founders say they remain in Budapest for personal and practical reasons: both live here and find strong engineering talent locally. The R&D and significant parts of operations are run from the Budapest office. They argue Hungary already has substantial domain knowledge in tree management, and with appropriate funding that know‑how could be turned into European‑leading practice.
Politics and budget choices
The founders note that lack of funding is often framed as "no money," but in many cases budgets are simply allocated elsewhere. They cite New York as an example: the city’s annual budget was described as USD 128 billion, while the greehill subproject would have cost around USD 4 million — showing that prioritization, not absolute absence of resources, is often the issue.
They also point out that repeated, systematic measurement allows cities to demonstrate the effectiveness of investments (for instance increasing canopy cover or improving average tree vitality), which helps justify public spending.
Lessons from other cities
In Singapore, crews first review trees virtually using the digital system and only go into the field when needed; extensive sensor networks and diverse, climate‑appropriate plantings bolster resilience. The founders say Europe has considerable expertise, but Singapore combines technical knowledge with sustained funding, enabling practice to match ambition.
Conclusion
The greehill case shows that combining LiDAR surveys with robust, task‑specific AI can rapidly produce operational tree cadastres that enable proactive urban tree management. According to the founders, the limiting factor in many places is political will and budgetary prioritization: without timely investment, climate change could render parts of urban environments far harder to live in within a few years.
Related event
Portfolio Sustainable World 2026, a sector‑neutral sustainability conference, will take place on 8 September 2026 and will cover green economy developments and operational interventions.
(Interview material: Goertz Gabor and Fekete Gyula; translations and editing based on the original Hungarian interview.)
