HKUST GREAT Smart Cities Institute Unveils ‘Streetsights’: A City-Scale AI Breakthrough Achieving 90%+ Taxi Demand Prediction Accuracy

In partnership with DASH, this landmark industry-academic collaboration replaces guesswork with real-time precision to boost driver earnings and urban efficiency. 

[9 July 2026] The GREAT Smart Cities Institute (GSCI) at the Hong Kong University of Science and Technology (HKUST) announced the results of Streetsights, a landmark smart mobility collaboration, at a press conference on 9 July 2026, together with industry partner Suntec Innovations Limited (Suntec). This locally developed, AI-powered demand forecasting system represents a major step forward in driving operational efficiency across Hong Kong's street-hail taxi network.

alt="A group photo of Jason Ma, the co-founder of DASH, Prof. Hong K. Lo, Director of GSCI and Mr Michael Chan, co-founder of DASH holding event signs at a press conference. The background is a large blue LED screen featuring a faint city skyline graphic, with white text that reads "GREAT Smart Cities Institute x DASH" and a line of traditional Chinese text below it. The man on the left wears a blue patterned blazer over a white shirt and holds a white "DASH" sign. Professor Hong K. Lo
(From left) Jason Ma, Co-founder of DASH; Prof. Hong K. Lo, JP, Director of GSCI; and Michael Chan, Co-founder of DASH, at the Streetsights joint press conference.

Turning Academic Excellence into Real-World Mobility Solutions

While street-hail taxis serve as a crucial pillar of Hong Kong’s public transit network for first- and last-mile connectivity, predicting where and when passengers will need a street-hail taxi has long been an industry-wide guessing game. Historically, a lack of aggregated data has forced operators and drivers to rely on intuition and personal experience to navigate shifting market demands.

Addressing this challenge head-on, the GSCI research team led by our Director, Chair Professor Hong K. LO, JP, and Research Engineer Dr. Zheng LIANG, developed a powerful AI model which leverages operation data from at least 2,000 street-hail taxis. This data, including the location and the occupancy status of taxis serving all districts in Hong Kong, is uploaded to the cloud every 5 seconds, to form a valuable big data pool for precise AI inferencing.
The Streetsights system automatically refreshes every 15 minutes, providing real-time demand predictions and taxi supply updates to help drivers navigate the city more efficiently. By turning raw data into a sophisticated demand forecasting system, the team successfully mapped shifting city-wide passenger demand and vehicle supply in real time.

(From left) Dr. Zheng Liang, Research Engineer of GSCI; Jason Ma, Co-founder of DASH; Prof. Hong K. Lo, Director of GSCI; and Dr. Weigang Lu, Research Team Member,
(From left) Dr. Zheng Liang, Research Engineer of GSCI; Jason Ma, Co-founder of DASH; Prof. Hong K. Lo, Director of GSCI; and Dr. Weigang Lu, Research Team Member,

Tangible Impact: Smarter Rides and Higher Earnings

By dividing Hong Kong into 350 designated zones (each spanning 3.75 sq km), the forecasting system predicts pickup demand with an outstanding accuracy rate of over 90 percent.

Projections suggest that drivers could complete 31.2 percent more orders and secure a 15.8 percent boost in daily fares by using the system. For passengers, this translates directly into a more reliable commute, significantly cutting down frustrating wait times during peak hours.

Prof. Hong K. Lo, Director of GSCI, presenting the Streetsights smart mobility breakthrough at the joint press conference.
Prof. Hong K. Lo, Director of GSCI, presenting the Streetsights smart mobility breakthrough at the joint press conference.

Paving the Way for Hong Kong’s Smart Mobility Future

GSCI’s latest breakthrough serves as a premier homegrown proof of concept, embodying our ‘GREAT’ framework to deliver a more Green, Resilient, Empowering, Adaptable, and Transformative mobility network. 

Established in 2018 alongside the Smart Mobility Research Cluster, GSCI leverages micro-simulation, big data, and AI to make urban travel safer, more comfortable, and highly efficient within existing city capacities. By partnering with government and industry, the cluster applies these tools to tackle real-world transit challenges. Notably, its cutting-edge forecasting models provide precise, large-scale optimization for Hong Kong, with practical applications already integrated into traffic engineering, urban planning, daily operations, and mass transit incident management.

Genuine partnerships of this kind pair world-class research with real-world industry data, proving that academic insights can seamlessly bridge the gap from the lab to the street. By modeling city-scale dynamics, GSCI has created a robust foundation for Hong Kong’s smart mobility future. 

Stay Connected

The GSCI research team for the Streetsights project is led by Prof. Hong K. LO and Dr. Zheng LIANG, and supported by team members Dr. Weigang LU, Mr. Helbert PAAT, and Ms. Trista LAI. To stay updated on our latest research milestones and event highlights, follow us on LinkedIn and Instagram. 

Back