Language Intelligence for Safer, Smarter Autonomous Mobility | Microsoft Fabric Thursdays Expert Series - 2025

 

 

Details Here!

RSVP on Meetup to Attend: 

meetup.com/up-powerbiclub/events/310373138

 

YouTube:

https://youtube.com/live/1Mc4mvrJD3o?feature=share

 

About Speaker

Rajani Acharya

University of Southern California | Engineering Manager Perception @ Luminar Technologies at Palto Alto, California, United States

She is an accomplished Engineering Manager with over 13 years of experience leading advanced machine learning and perception systems for autonomous vehicles, robotics, and ADAS with multiple patents granted in camera-radar fusion and autonomous system design. Rajani is known for her strategic leadership, strong cross-functional collaboration, and a relentless drive toward safety, scalability, and technical excellence in AI-driven systems.

 

Session Details for Aug 7, 2025: 

Large Language Models (LLMs) are revolutionizing the autonomous mobility landscape — enabling deeper contextual understanding, flexible decision-making, and more natural human-machine interactions. Join us for an exclusive deep dive into real-world LLM deployments that are breaking away from traditional AV (Autonomous Vehicle) pipelines.

In this session, we’ll explore:

• Waymo’s EMMA system: Achieved 16% fewer perception errors and 24% fewer disengagements using a unified transformer with cross-modal attention.
• DeepSeek R1 (China): A reasoning-first driving model adopted by BYD and others, designed to align with local driving norms.
• DriveMLM Framework: Boosts command interpretation accuracy by 87% and improves user trust by 34%.

We’ll also cover key advances such as:

• Up to 18% better rare object classification
• 23% reduction in false positives via contextual LLM filtering
• Technical challenges like high computational demands (>100 TOPS), real-time constraints, and regulatory considerations for probabilistic behaviors

This session will also introduce emerging solutions including:

• Hybrid neuro-symbolic architectures
• SOTIF (Safety of the Intended Functionality) compliance
• Edge AI optimizations

Our expert speaker, who holds 2+ patents in the field, brings deep technical insight and hands-on experience in deploying LLMs in safety-critical environments.

 

Be Ready to Engage:

• Bring a notepad – this is a content-rich session!
• Submit your questions beforehand via Comments.
• Ask questions live via comments or join the stage with camera/mic access – just reach out to us in advance if you’d like to participate on screen.

Don’t miss this opportunity to explore the future of autonomous driving through the lens of AI and language intelligence.

Email: powerbiclub-up@hotmail.com