Date: Thursday December 4^th, 2025

Time: 4:00pm PST

Location: MGM

 

Join us to explore how Edge AI can enhance physical systems with improved speed, efficiency, and cost savings. We'll examine a real-world manufacturing defect detection use case and demonstrate how to integrate AWS SageMaker with Qualcomm Dragonwing™ processors and AWS IoT Greengrass. The session will also cover transforming legacy devices by adding intelligence and cloud connectivity using Qualcomm AI Hub and Edge Impulse with AWS solutions. This will showcase practical applications and triggered actions for Edge AI implementations.

 

Speaker: Krishna Sridhar, Vice President Engineering, Qualcomm Technologies, Inc.
Ashvin Roharia, Senior Engineer, Solutions Engineering – Post-Sales Focus, Qualcomm Technologies, Inc.

Date: Monday, December 1^st, 2025

Time: 2:30 PM - 3:30 PM 

Location: Mandalay Bay, Level 2 South, Lagoon G

 

This talk discusses delivering reliable GenAI features on millions of devices by benchmarking AI mobile apps on real iOS and Android devices. It highlights the differences between emulators/simulators and real devices and showcases Generative AI-driven test platforms. Qualcomm AI Hub optimizes, compiles, and tests models across various targets with physical device validation using AWS Device Farm. The talk includes a case study from the Amazon Retail team, demonstrating accelerated testing and on-device AI capabilities.

 

Speaker: Krishna Sridhar, VP Engineering, Qualcomm, Nikhil Dabhade, Principal Solutions Engineer, Amazon Web Services, Carlos Arguelles, Senior Principal Engineer, Amazon

This demo illustrates how Qualcomm AI Hub streamlines the end-to-end developer journey, featuring sample apps in Large Language Models (LLMs) and Object Recognition. These applications are deployed across a diverse set of devices—including a smartphone, laptop, and the RB3 Gen 2 development kit—demonstrating the flexibility and scalability of Qualcomm’s platforms. The models showcased in the demo are trained using AWS SageMaker, highlighting seamless cloud-to-edge integration.

Using a pharmaceutical manufacturing use case, see how edge AI cameras powered by the Qualcomm^® Dragonwing™ QCS6490 processor in the Advantech AOM-2721 can be combined with Edge Impulse, AWS Greengrass for computer vision based AI to spot defects on the production line automatically – like cracks or misprints on a pill. This setup can speed quality checks, reduce waste and keep things efficient and secure since all the data stays onsite (at the edge). This solution can be applied to many industrial and manufacturing scenarios, such as detecting pipe corrosion, detecting defects in electronic control systems and touchscreen interfaces, mislabeled or damaged food packaging and more.

This demo introduces the concept of a physical twin in the cloud, enabling remote access to the Qualcomm SA8650P SoC through a secure portal for developing and testing automotive applications. Developers can simulate real-world scenarios and iterate rapidly without needing physical hardware on-site, streamlining the development cycle. Complementing this, the demo also features the QCR100 platform and the cloud-based Data Simulation Factory (DSF), which is jointly optimized with AWS to support advanced data reprocessing and validation workflows. This integration empowers automotive developers to accelerate innovation by combining high-performance edge computing with scalable cloud-based simulation and validation capabilities.

This interactive demo showcases how computer vision can enhance future smart cities by improving road safety and reducing vulnerable road user accidents, which number in the millions annually according to the World Health Organization. Using Edge Impulse enabled AI and the Arduino Uno Q powered by the Qualcomm Dragonwing™ QRB2210 processor, the demo performs real-time vehicle detection, object tracking, and crash event recognition all running efficiently at the edge. Participants race miniature slot cars around a track while a camera monitors lap times and displays a leaderboard of the fastest racers, offering a fun and engaging way to experience the potential of edge AI in urban mobility and safety applications.