Issue link: https://maltatoday.uberflip.com/i/1547431
MALTA cannot build its way out of congestion. It can get much better at using the roads it already has, and that starts with seeing problems early. Picture a weekday morning. A car breaks down in the left lane of the Marsa-Hamrun bypass, the queue grows back towards Qormi, and buses on several routes start running late. Most of the people stuck in it only find out what happened when they finally crawl past the haz- ard lights. Nobody did anything wrong. The information simply arrived too late to be useful. For the past 18 months, our team at Digital Brain has worked to close the gap between some- thing happening on the road and someone being able to act on it. The project, Digital Traf- fic Brain, was financed by Xjen- za Malta through the FUSION R&I Technology Development Programme LITE. It set out to answer a practical question: How much can a small coun- try improve its traffic without building new roads or installing new sensors? The first thing the data taught us is that Malta's traffic is far more predictable than it feels. Every weekday morning, cars flow into Valletta, Birkirka- ra, Qormi and San Ġwann. By midday the pressure moves to commercial areas such as Sliema and Pembroke, and at night to Paceville. Rain changes everything—drivers leave big- ger gaps, speeds drop, and the whole network slows at once. When patterns are this regular, a system can learn them and warn people before the queue forms, not after. The second lesson is that bus- es suffer most. Public transport in Malta is free, yet a bus caught behind a breakdown is just as stuck as every car around it, and each delay pushes more people back into their cars. So, we mapped every bus stop and route onto the same road net- work as the live traffic data. When an incident is reported, the platform shows which bus routes are affected and where delays are building. What came out of the project is a working platform. It pulls live incident and congestion reports from sources such as Waze and TomTom every few minutes, ties each one to the exact road segments affected, and keeps a history of where disruptions keep surfacing. It also uses models that predict delays on individual roads and computer vision that recog- nises lanes and flags a vehicle stopped in a main lane, so au- thorities can be alerted straight away. Commuters can plan journeys that combine walking, buses, ferries, and driving based on what the roads look like now rather than a timetable. None of this requires digging up a road, and in a country with little space left to build and where every major road project takes a substantial amount of time, that matters. Better in- formation will not empty our roads. However, it can shorten the time between a problem ap- pearing and someone respond- ing; whether it's a traffic officer being sent to a breakdown, a bus operator rerouting a ser- vice, or a driver choosing to leave 10 minutes later. We also built the platform so that it is not tied to Malta. Everything specific to a city, from its boundaries to its ferry routes, sits in a configuration file, so a new city can be added without rewriting the software. When we showed the system at the Smart City Expo World Congress in Barcelona, munic- ipalities from Europe, the Mid- dle East and Latin America were impressed at how quickly it could run on their own streets. A solution developed for Mal- ta's very particular problems turns out to travel well. Projects like this rarely start in a large company with a large budget. They start with a small team, a hard problem and someone willing to back an idea before it is proven. Xjen- za Malta's support through the FUSION programme gave us time to do the research proper- ly, test it on real Maltese data, and move it from a prototype to a system ready for deployment. The next step is to put it in the hands of the people who man- age our roads and buses every day. Traffic will remain one of the hardest problems on this is- land, but predicting problems before they happen is far better than reacting to them after the fact. The Digital Traffic Brain project was financed by Xjen- za Malta through the FUSION R&I Technology Development Programme LITE. 9 maltatoday | SUNDAY • 1 NOVEMBER 2026 OPINION What 18 months of traffic data taught us Alexiei Dingli & Luca Bondin Dingli is a professor in AI and Bondin is an expert in affective computing and synthetic data Traffic will remain one of the hardest problems on this island, but predicting problems before they happen is far better than reacting to them after the fact