eco-driving

IntersectionZoo: Eco-driving for Benchmarking Multi-Agent Contextual Reinforcement Learning

Open code of my published paper: Health-considered energy management strategy for fuel cell hybrid electric vehicle based on improved soft actor critic algorithm adopted with Beta policy

Open code of my published paper: Ecological Driving Framework of Hybrid Electric Vehicle Based on Heterogeneous Multi-Agent Deep Reinforcement Learning

[ECC 2022] Codebase for the paper titled "Learning Eco-Driving Strategies at Signalized Intersections".

This repository contains codes and datasets in correspondence to my bachelor thesis for computer science at Vrije Universiteit Amsterdam.

Open code of my published paper: Eco-Driving Framework for Hybrid Electric Vehicles in Multi-Lane Scenarios by Using Deep Reinforcement Learning Methods

Code for 'Mitigating Metropolitan Carbon Emissions with Eco-driving at Scale' published at Transportation Research Part C: Emerging Technologies (TR-C) 2025.

This is an app that demonstrates using of the Telematics SDK and walks you through the integration. The SDK tracks user location and driving behavior such as speeding, cornering, braking, distracted driving, and other parameters.

Does an LSTM traffic-speed forecast make EV eco-cruise control more efficient? LSTM-PID and LSTM-MPC on the EPA UDDS/HWFET/US06 cycles: a perfect 2s preview saves 0.7–4.3% energy at equal tracking; the LSTM recovers little of it.

Sparkilo — find the cheapest fuel & EV charging near you across 17 countries. Real-time official prices, route/best-stop planning, price alerts, OBD2 + GPS trip logbook, consumption & cost tracking. Open-source, ad-free, no account, privacy-first.

Featherfoot — score how smoothly you drive from your phone's motion sensors alone. A leaf that wilts when you're rough and a braking/cornering/acceleration traffic light to share. Android, offline, no GPS.

Personalising an eco-driving HUD with human-in-the-loop multi-objective Bayesian optimisation in VR (UCL MSc thesis study)

Greenwave project website.

AI-powered eco-driving analysis platform using React, FastAPI, MongoDB, and Ollama LLM.

Supplementary bibliographic data for an RL-based eco-driving review covering 1991–2026. Search coverage through 16 August 2026.

Explainable driving-behaviour classifier (XGBoost + SHAP + FastAPI): labels 10 s of GPS as smooth/moderate/aggressive, weighted-F1 0.946. Part of EcoRoute, Monash FYP 2026.

openHAB 5 binding for the Traccar GPS server: real-time position via webhooks, geofence events, Nominatim reverse geocoding, distance tracking, commands to the tracker, eco-driving, Teltonika CAN adapter (ALL-CAN300 / LV-CAN200) data, OBD-II and BLE beacon tracking. Cloud or self-hosted Traccar.

EcoRoute mobile app (Expo/React Native): eco-route planning, turn-by-turn navigation, real-time driving coaching and per-trip CO₂ from phone GPS. Monash FYP 2026.