Introduction
What DriveX is building
DriveX is UCLA Mobility Lab’s public platform for advancing physical AI in intelligent transportation through recurring workshops, open challenges, research assets, and seminars.
The initiative connects frontier methods in reasoning, world models, and embodied autonomy with deployment-oriented evaluation, enabling sustained collaboration across academia and industry.
Purpose
Mission
The mission of DriveX is to advance the next generation of intelligent transportation and autonomous systems by building a research and community platform around physical AI for real-world mobility.
DriveX aims to support progress not only in core AI capabilities such as multimodal perception, embodied reasoning, world models, and 3D/4D scene understanding, but also in the deployment-oriented challenges that define transportation systems in practice, including cooperative autonomy, infrastructure-assisted intelligence, robustness, evaluation, and real-world integration.
By bringing together research, benchmarking, open assets, and recurring community programs, DriveX seeks to accelerate both scientific progress and practical impact.
Definition
What DriveX is
DriveX is a broad initiative built around the idea that the future of intelligent transportation will require deeper integration between AI, autonomy, infrastructure, and real-world deployment.
At the initiative level, DriveX provides a shared umbrella for a growing ecosystem of activities, including:
- recurring workshops and tutorials across leading AI, computer vision, robotics, and intelligent transportation venues;
- challenge programs and benchmark tracks that support measurable research progress;
- datasets, tools, and technical resources that enable open research and evaluation;
- seminars and invited talks that connect academia, industry, and emerging researchers;
- and broader community-building efforts that strengthen collaboration across institutions and disciplines.
DriveX is designed to provide continuity across years, venues, and technical themes while remaining open to collaboration from partner institutions and program organizers.
Technical breadth
Research scope
DriveX spans a broad set of research areas at the intersection of physical AI and intelligent transportation. These include, but are not limited to:
- embodied reasoning and decision-making for autonomous systems;
- world models and predictive environment representations;
- multimodal perception and scene understanding;
- 3D/4D scene intelligence and spatial representations;
- cooperative perception, V2X, and infrastructure-assisted autonomy;
- robustness, generalization, and evaluation for real-world deployment;
- digital twins, simulation, and deployment-oriented testing;
- and broader systems integration for safe and effective intelligent transportation.
This breadth is intentional. DriveX is designed as a broad umbrella initiative, while allowing each year’s programs to emphasize the topics most relevant to host venues and active partners.
DriveX is UCLA-led—partners and co-organizers join as contributors to a single open platform, not separate silos.
How it fits together
How DriveX works
DriveX is organized as a unified research platform with multiple public program tracks under one UCLA Mobility Lab framework.
Faculty leadership sets scientific direction, while partner labs, students, and industry teams contribute through workshops, challenges, assets, and seminars.
Current program pillars include:
- DriveX Workshop Series — recurring forums across leading AI, vision, robotics, and intelligent transportation venues.
- DriveX tutorials — instructional tracks aligned with the same annual themes as the workshop series.
- DriveX Grand Challenge — benchmark tracks spanning reasoning, multimodal perception, cooperation, robustness, and real-world autonomy.
- DriveX Research Assets — datasets, baselines, challenge kits, and tooling released for open reuse.
- Seminars & community — invited talks and exchanges connecting students, faculty, and partners.
Rationale
Why DriveX
The future of intelligent transportation will not be shaped by isolated technical advances alone. It will require stronger connections among frontier AI research, embodied systems, infrastructure, evaluation, and real-world deployment.
DriveX was created to help build those connections.
By combining scientific ambition with recurring public programs, open technical assets, and a broader ecosystem perspective, DriveX aims to serve as a durable platform for advancing physical AI in one of the most important and complex real-world domains: mobility and transportation.
People
Team & collaborators
DriveX grows through UCLA leadership plus co-organizers and partner institutions across academia and industry.
The People page introduces the faculty lead and core program stewards you’re most likely to meet through our public activities. Browse the full Mobility Lab roster on the lab people page, and see institutional partners on Partners.
Get involved through workshops, Grand Challenge tracks, research assets, or seminars based on your team’s interests.