Hong Kong Industrialist Jan-Mar 2021

LOCAL DISCOVERY | 本地科研 Intelligent Transportation System In early 2010s, the electronic toll collection company Autotoll Limited and PolyU worked out an innovative solution for the government to detect various traffic data for estimation of current travel times on major routes in the city. Prof Lam and his team thus developed an ITS that integrates limited traffic data from different sources to estimate the real-time traffic condition with an accuracy of 95%. The results are displayed and updated once every two minutes on traffic sign boards and the website of the Transport Department as colour- coded speed map and journey time so that motorists may then take alternative routes to avoid congestion. From 2018 to 2020, different types of traffic detectors were installed on strategic routes of Hong Kong to further enhance coverage and accuracy of data capture. Prof Lam and his team have been developing a sophisticated solution method to help integrate different data from multiple sources, including real-time data from different traffic detectors and short- term traffic prediction results, with use of both advanced machine-learning algorithms and complex traffic simulation models, in order to further enhance the real-time traffic prediction on major roads in Hong Kong. It is expected that the predicted journey time information can be refreshed every minute. HKeTransport Public transport is the only or preferred way of local travel for the majority of commuters in Hong Kong. Despite the ubiquity of trip planning apps and websites available these days, there was none around the turn of the millennium. In 2002, PolyU was among the first to provide a public transport route search service for its students and staff, known as EasyGo. Meanwhile, the Transport Department was looking to develop a multi-modal public 智能交通系統 2010年代初，政府計劃為香港的主要路 線提供行車時間估算服務。理大遂與電子道 路收費服務供應商快易通有限公司攜手合 作，成功制定一個創新的算法。林教授的團 隊研發相關的智能交通系統，通過整合不同 來源的有限交通數據，估算實時交通狀況， 準確度高達95%。所得結果除在交通路牌上 每兩分鐘更新一次以外，還會顯示於運輸署 網頁，在地圖上以顏色區分車速，輔以預計 實時行車時間，方便駕駛者因應塞車情況作 出繞道的決定。 由2018到2020年間，當局在香港主要 幹線上安裝多種交通檢測器，從而進一步提 升覆蓋範圍和數據採集的準確度。因此，林 教授和他的團隊現正研發一個精密的算法， 以整合來自不同來源的數據，包括不同交通 檢測器的實時數據和短期交通預測結果，並 運用先進的機器學習算法和綜合交通模擬模 型，進一步預測本港主要道路的實時交通情 況，預期可每分鐘更新資訊一次。 Intelligent Transportation System and HKeTransport Platform The interface of HKeTransport 香港乘車易的界面 58 | 1-3/2021