Welcome.AI
    Skip to content
    AI Agents

    Wayve’s AI Driver Enhances Deployment Efficiency and Market Reach

    Wayve's AI Driver is set to revolutionize the autonomous vehicle industry by learning in real-time from diverse environments, eliminating the need for extensive pre-mapping. With backing from key automotive players, this technology promises faster deployment of robotaxis and enhanced consumer vehicle capabilities.

    coreweave.com•November 21, 2026•2 min read

    Key Facts

    • Wayve's AI Driver reduces deployment time by eliminating pre-mapping, enhancing competitive edge.
    • Training on NVIDIA Blackwell GPUs allows rapid scaling, crucial for meeting growing demand.
    • Access to diverse real-world data from 70+ countries strengthens model performance and safety.
    • Partnerships with major automakers enable Wayve to tap into a market of over 10 million vehicles annually.
    • Centralized learning from varied deployments creates a data flywheel, driving continuous improvement.

    Summary

    Summary

    Wayve, a company specializing in AI-driven autonomous vehicles, faced challenges in efficiently training its AI Driver model using vast datasets from diverse environments. By partnering with CoreWeave to utilize NVIDIA Blackwell GPUs, Wayve significantly accelerated its deployment of the AI Driver, resulting in successful launches of its robotaxi service in London and planned expansions to Tokyo.

    Background

    Wayve, founded in 2017, operates in the autonomous vehicle industry, focusing on developing an AI Driver that learns to navigate from real-world data rather than relying on pre-built maps. Prior to deploying its AI Driver on CoreWeave, Wayve managed a complex training process that required extensive computational resources and faced limitations in scalability and reliability.

    Challenge

    Wayve needed to overcome the bottleneck of training its AI Driver model efficiently while managing large-scale datasets and ensuring that the infrastructure could support long training runs necessary for safe and effective deployment.

    Solution

    Wayve signed a multi-year agreement with CoreWeave to utilize its AI cloud infrastructure, specifically leveraging NVIDIA Blackwell GPUs. This partnership allowed Wayve to scale its training capacity flexibly, ensuring that GPUs remained well-fed with data during training sessions. The deployment pathway included multi-month pre-training, large-scale offline evaluation, and on-road validation, all supported by CoreWeave's robust infrastructure.

    Results

    Wayve successfully transitioned its AI Driver from research to production, launching a supervised robotaxi service in London and planning to expand to Tokyo. The model has demonstrated its capability by navigating challenging environments, including a typhoon in Tokyo and the Arc de Triomphe in Paris, without prior specific training on those roads. The ongoing data collection from these deployments creates a feedback loop that continuously improves the model.

    Key Insights

    The partnership with CoreWeave highlights the importance of having a scalable and reliable cloud infrastructure for AI training, particularly in the autonomous vehicle sector. Access to the latest GPU technology enables faster iterations and model improvements, which is crucial for maintaining a competitive edge.

    Customer Testimonial

    "Our AI Driver gets better by running this data flywheel: gathering real-world driving experience, training on it, evaluating it, and deploying only what outperforms what is already on the road," said Silvius Rus, SVP of Engineering at Wayve.

    Entities Mentioned

    Companies

    Wayve
    CoreWeave
    NVIDIA
    Uber
    Nissan
    Stellantis
    Mercedes-Benz

    Products

    AI Driver
    NVIDIA Blackwell GPUs

    Technologies

    AI
    robotaxis
    foundation models
    reinforcement learning

    People

    Alex Kendall
    Silvius Rus

    Key Concepts

    end-to-end foundation model
    real-world data
    robotaxi deployment
    AI cloud infrastructure
    data flywheel
    multi-year agreement
    large-scale sensor data
    GPU training

    Definitions

    AI Driver
    An end-to-end foundation model developed by Wayve that learns to drive using real-world data.
    data flywheel
    A process where increased data collection improves model performance, leading to more deployments and further data collection.
    reinforcement learning
    A type of machine learning where an agent learns to make decisions by receiving rewards or penalties.
    robotaxi
    An autonomous vehicle designed to transport passengers without a human driver.
    foundation model
    A large-scale AI model that can be adapted for various tasks with minimal additional training.

    Use Cases

    • →Deployment of robotaxis in new cities
    • →Consumer vehicles equipped with AI Driver
    • →Training AI models on large-scale datasets
    • →Real-time driving experience evaluation
    • →Centralized learning from diverse applications
    • →Commercial launches of autonomous driving technology

    Frequently Asked Questions

    What is the AI Driver?

    The AI Driver is an advanced model developed by Wayve that learns to drive using real-world data rather than traditional mapping methods. It aims to improve the deployment of autonomous vehicles.

    How does Wayve's training process differ from traditional methods?

    Wayve's training process relies on real-world data and a unique workload that differs from training language models. It involves multiple phases, including pre-training, evaluation, and on-road validation.

    What role does CoreWeave play in Wayve's operations?

    CoreWeave provides the AI cloud infrastructure that Wayve uses to train its models. This includes access to NVIDIA GPUs and the ability to scale resources flexibly as training demands change.

    What are the benefits of using NVIDIA Blackwell GPUs?

    NVIDIA Blackwell GPUs allow Wayve to train larger models more efficiently and keep up with the growing demands of AI training. They also enable faster iterations and improvements in model performance.

    What is the significance of the data flywheel for Wayve?

    The data flywheel is crucial for Wayve as it enhances the AI Driver's performance through continuous data collection and learning. More data leads to better models, which in turn drives further deployments and data acquisition.

    Where AI Leaders Stay Informed

    The latest AI intelligence, case studies, and research — delivered to your inbox every week.

    Free to read. Unsubscribe anytime.