How AI Powers Gaode’s Route Planning and Navigation: Inside the Tech

The Gaode (Amap) Technology Open Day has concluded, and a 17‑minute video replay of a presentation by Cui Hengbin (nickname “QianTong”) from the Machine Learning R&D team is now available.

Using a typical user travel scenario, the speaker demonstrates how AI technologies influence the journey before departure, during travel, and after arrival to enhance user experience. Key areas include road‑condition prediction, navigation path planning, ETA (estimated time of arrival) calculation, and dynamic traffic‑event mining, all powered by a suite of innovative machine‑learning techniques.

The talk references two notable research papers: “Hybrid Spatio‑Temporal Graph Convolutional Network: A More Accurate Spatio‑Temporal Prediction Model” (KDD 2020) – https://mp.weixin.qq.com/s?__biz=Mzg4MzIwMDM5Ng==∣=2247485057&idx=1&sn=f5f5dccaf41a00c76d62118bf8a42aa7 and “LRC‑BERT: Contrastive Learning for Latent Semantic Knowledge Distillation” (AAAI 2021) – https://mp.weixin.qq.com/s?__biz=Mzg4MzIwMDM5Ng==∣=2247485799&idx=2&sn=c2e938e3a93e97d47c55d1daab72aee1.

The video runs for 17 minutes, including a Q&A session. For further technical exchange, the speaker can be contacted at [email protected].