NeuralMOVES: A lightweight and microscopic vehicle emission estimation model based on reverse engineering and surrogate learning
Edgar Ramirez-Sanchez, Catherine Tang, Yaosheng Xu, Nrithya Renganathan, Vindula Jayawardana, Zhengbing He, Cathy Wu · 2025
Read NeuralMOVES: A lightweight and microscopic vehicle emission estimation model based on reverse engineering and surrogate learning by Edgar Ramirez-Sanchez, Catherine Tang, Yaosheng Xu, Nrithya Renganathan, Vindula Jayawardana, Zhengbing He, Cathy Wu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-04417.txt.
Description
Edgar Ramirez-Sanchez and co-authors built NeuralMOVES in 2025, a fast stand-in for the standard vehicle emissions simulator. Transport is a major source of greenhouse gases, and planners need emission numbers to guide cuts. The certified standard tool, MOVES, is heavy: complex to run, slow, and unsuited to second-by-second estimates in live traffic.
Working backward from that tool's behavior, the team trained neural surrogates that mimic it at microscopic scale. Across more than two million test scenarios with varied driving traces and conditions, the surrogate's average error against the standard stayed near six percent, in the figure reported. Readers get emission estimates at simulation speed: a heavyweight model distilled into a lightweight one that traffic tools can actually call.
Catalog record
| CREATOR | Edgar Ramirez-Sanchez, Catherine Tang, Yaosheng Xu, Nrithya Renganathan, Vindula Jayawardana, Zhengbing He, Cathy Wu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-04417