Indonesia Opens Its First University AI Centre to Tackle TB, Floods and Farming
A new lab in Yogyakarta is giving Indonesian researchers access to serious computing power for the first time, with three early projects targeting the country's most urgent health and safety problems.
Key points
- Indonesia's Ministry of Communication and Digital Affairs, Indosat, NVIDIA and Universitas Gadjah Mada launched the country's first university-based AI technology centre in Yogyakarta in 2025.
- The centre runs on Indosat's GPU Merdeka platform, giving students and researchers access to enterprise-grade AI computing that was previously out of reach.
- Three funded research projects target tuberculosis screening, precision farming, and disaster early-warning systems.
- Indonesia records more than one million new tuberculosis cases every year; one project aims to detect the disease from a breath sample, no specialist needed.
- The initiative is part of Indonesia's national AI Centre of Excellence programme, designed to shift the country from AI consumer to AI producer.
Indonesia is the fourth most populous country on earth. It has no shortage of researchers or hard problems to solve. What it has lacked, until now, is the raw computing muscle to turn ideas into working technology.
That changes, at least a little, with a centre that opened this week in Yogyakarta.
The UGM Indosat NVIDIA AI Technology Centre brings together four partners: the government's Ministry of Communication and Digital Affairs, telecoms operator Indosat Ooredoo Hutchison, chip company NVIDIA, and Universitas Gadjah Mada (UGM), one of Indonesia's most respected public universities. As first reported by NVIDIA, the centre is the first of its kind based inside an Indonesian university.
What will researchers actually be able to do?
For the first time, UGM's students and scientists can access enterprise-grade GPUs (graphics processing units, the specialised chips that do the heavy number-crunching AI needs) without travelling abroad or paying prohibitive cloud bills. The computing comes through Indosat's GPU Merdeka platform, a sovereign GPU-as-a-service system, meaning the computing power sits inside Indonesia rather than being rented from foreign data centres. Researchers also get access to NVIDIA's Nemotron open models (pre-built AI systems that teams can customise rather than build from scratch) and technical mentorship.
What are the first three projects?
The centre launched with three concrete research programmes, each aimed at a problem that costs Indonesian lives or livelihoods.
Tuberculosis screening. Indonesia records over one million new TB cases every year. TB is curable, but it kills when it goes undetected, particularly in rural areas where specialist doctors and expensive lab equipment are absent. A team from UGM's Faculty of Medicine is developing eNose-TB, an electronic nose that analyses a patient's breath sample for signs of infection. The goal is a fast, affordable screening tool that works in remote clinics and villages, no laboratory required.
Precision farming. Agriculture employs close to 30 percent of Indonesia's workforce. SmartAgri combines satellite imagery, sensor readings and local farming knowledge using multimodal AI (systems that process several types of data at once) to give smallholder farmers precise guidance on when and how to irrigate their fields.
Disaster early-warning. Indonesia sits on the Pacific Ring of Fire and faces some of the highest natural disaster risk on the planet. Tech4Disaster will use AI to process satellite and sensor data at speed, giving communities and emergency responders earlier warning before floods, earthquakes or volcanic events hit.
| Project | Problem targeted | Key method |
|---|---|---|
| eNose-TB | 1 million+ annual TB cases | Breath-sample AI screening |
| SmartAgri | Rural farming productivity | Satellite and sensor data fusion |
| Tech4Disaster | Earthquake and flood risk | Geospatial AI on accelerated computing |
What does this mean for ordinary Indonesians?
None of these tools are in clinics or farms yet. The centre is a research facility, not a product launch. But the direction is clear: if eNose-TB works, a village health worker could screen patients without a specialist. If SmartAgri delivers, a smallholder farmer gets the kind of precision advice once available only to large commercial operations.
The government framed the launch explicitly as an attempt to move Indonesia from a country that uses AI built elsewhere to one that builds its own.
Common questions
Is this centre open to all Indonesian universities, or just UGM?
For now, UGM hosts and leads the centre, but the stated goal is to connect researchers, startups and innovators across Indonesia to the platform over time.
How far away are these projects from real patients or farmers?
All three are in early research stages. eNose-TB is the furthest along conceptually, but any medical screening tool must pass clinical trials and regulatory approval before it reaches a clinic, a process that typically takes several years.
