Elm’s research in autonomous mobility focuses on advancing and testing next-generation self-driving software through the development of artificial intelligence systems capable of perceiving their surroundings and making safe, real-time decisions. Our experimental platforms feature vehicles equipped with cutting-edge technologies — including sensors, cameras, radar, LiDAR (Light Detection and Ranging), GPS, and advanced algorithms for perception, decision-making, and motion control — to enable the future of intelligent and autonomous mobility solutions.
Research Areas
Digital Mapping & Digital Twins
Multi-modal 3D Digital Mapping Systems
Development of accurate and scalable digital mapping systems for generating digital replicas by humans, robots, drones and autonomous vehicles.
HD and 3D Map Generation and Maintenance for Smart Cities and Autonomous Mobility
Scalable end-to-end software platforms for generation and management of city-scale HD and 3D maps using advanced technologies such as sensor fusion, SLAM and Gaussian Splatting to unlock smart city applications.
Sovereign AI-embedded Geospatial Platforms
Research on building sovereign geospatial platforms with embedded AI agents for automatic extraction of features, smart navigation and geospatial analytics.
Autonomous Mobility
Autonomous Vehicles for Last Mile Mobility
Development of use-case specific self-driving vehicle platforms (ex. shuttles, carts, cars, etc.) powered by the Elm Virtual Driver for passenger transportation in geofenced environments.
Innovative User Interfaces in Autonomous Vehicles
Analysis and design of novel user experiences for passengers in connected and autonomous vehicles, taking into consideration cultural nuances and linguistic preferences for Saudi region.
Learning based scene understanding and self-driving models
Leveraging and adapting models such as Vision Language Models for Saudi driving environments by creating specialized datasets, fine tuning methods, and fusion strategies to enhance reliability, interpretability and safety in decision making tasks.
- Published Research Papers
1. Vehicle-to-everything (V2X) in the autonomous vehicles domain – A technical review of communication, sensor, and AI technologies for road user safety
2. Real-time Instance Segmentation Models for Identification of Vehicle Parts