Webinar: From Perception to Action: Closing the Gap for Humanoid Deployment

Humanoid robots are advancing rapidly, but a fundamental challenge remains: turning perception and intelligence into fast, reliable physical action. In dynamic, unstructured environments, robots must continuously sense, interpret, decide and act - and today that loop is often still too slow and uncertain to match human performance.

This webinar will explore what it takes to close the perception-to-action gap, including advances in 3D and multimodal perception, AI models, simulation and synthetic data.

We'll examine why real-world training data alone cannot represent every environment, interaction and edge case humanoids will encounter, and why simulation grounded in real-world sensing will be critical to more representative training and validation.

The discussion will examine how bringing sensing, simulation, intelligence and action closer together can help move humanoids beyond impressive demonstrations toward scalable, productive deployment in the real world.

Date: Wednesday, October 14, 2026
Time: 12:00 P.M. EDT/9:00 A.M. PDT

Questions to be addressed in the program include: 

• Why humanoids still struggle to perform unstructured tasks at human speed

• How 3D and multimodal perception can tighten the loop between sensing, decision-making and action

• Why real-world training data alone cannot capture every environment and edge case

• How simulation and synthetic data can expand training and validation

• Why grounding simulation in real-world sensor data is critical to more representative performance

• What advances are still needed to move humanoids from controlled demos to scalable deployment

About the speakers

Alon Zehngut
Director of Product for Robotics, RealSense

Alon Zehngut is Director of Product for Robotics at RealSense, where he leads product strategy and go-to-market for advanced perception technologies powering the next generation of physical AI systems. He brings more than a decade of experience across robotics, autonomous systems, and defense, with expertise in visual perception, multi-sensor fusion, and deploying complex systems in real-world environments.

Paul Li
Head of R&D, LimX Dynamics

Paul Li is Head of R&D at LimX Dynamics, a leading embodied intelligence and humanoid robotics company, where he leads the company’s robotics research and development. His work focuses on advancing reliable, capable robotic systems and translating cutting-edge research into real-world applications.

Tim Culverhouse
Editorial Director, Robotics 24/7

Tim is the Editorial Director of Robotics247.com. His mission is to provide valuable information and insights to robotics professionals and decision-makers, and to help them solve business challenges. He is a creative, deadline-driven and detail-oriented storyteller.