Zebra Technologies Corporation · NASDAQ:ZBRA
Zebra Technologies — Technical Presentation (2023)
14 slides from Zebra Technologies's technical presentation, 2023
Slide 1: AI on the Edge
AI on the Edge Stuart Hubbard Senior Director Artificial Intelligence & Advanced Development 1
Slide 2: Agenda
Agenda AI on the Edge • Demo • Technical Benefits: Definition, Importance & Advantages • The role of AI in Edge Devices: Insights, Personalization, Optimization & Shrink prevention • Tools for Accelerating AI Application Development • Optimization & metrics for on-device • Examples • Future Trends •
Slide 3: Product Demos
Product Demos Running on-device “We expect 1.9 billion edge devices to ship with deep learning accelerators in 2025,” Linley Group principal analyst Linley Gwennap explained. Product Recognition – without the barcode More in the AR Talk LLM on TC58 ZEBRA TECHNOLOGIES
Slide 4: Technical Benefits
Technical Benefits • Low Latency and Real-time Processing: Improved user experience and reduced cloud costs / “AI on edge devices means complexities implementing AI capabilities directly • Reduced Bandwidth Usage & Cloud Costs: on local devices without constant Not relying on customer's network or s
Slide 5: The role of AI in Edge Devices
The role of AI in Edge Devices A few examples: • Real-time Insights: Better decision making, improved customer experience, optimized operational efficiencies, early identification of issues… • Personalization: Active learning from behavior and improved customer experience. • Optimization: Real-time
Slide 6: Tools for Accelerating AI Application Development
Tools for Accelerating AI Application Development ZEBRA TECHNOLOGIES 6
Slide 7: Enabling Partners & ISVs
Enabling Partners & ISVs Device Stack – SDK & more… Retail Analytics Loss Prevention Associate Enablement Task Management (Out of Stock) (Ticket Switching) (Voice co-pilot) (Intelligent Task Allocation) Customer ISV Software Stack Enabling Partners to collect data, train AI Models and deploy across
Slide 8: Zebra's AI Development Framework
Zebra’s AI Development Framework APPLICATION RETAIL INDUSTRIAL T&L HEALTHCARE EXAMPLES Associate Task Machine Machine / Health Last Mile Delivery Patient Tracking Retail Analytics Loss Prevention Enablement Management Vision Status & Task ISV & Data Collection Model Builder Model Library Deployment
Slide 9: End-to-end Pipeline
End-to-end Pipeline A mature AI Development Framework offers increased efficiency, enhanced model performance, scalability, robust governance, cost optimization, agility, and improved collaboration. These benefits collectively contribute to driving business growth, competitive advantage, and success
Slide 10: Optimization & metrics
Optimization & metrics Product Recognition • Optimized for Zebra devices, partnership with Google and Qualcomm TC58 (QC 6490) 200k product database (~80Mb) • State-of-the-art techniques – quantization, model networks, layers used for output. Example of performance: • ~46ms to localise and perform a
Slide 11: Examples
Examples Self Checkout – running on-device Build business rules to ”nudge” customer or alert staff • Own bag • Stacked Products • Missed Scan • Ticket Switch Fresh Produce Recognition Ticket Switching ZEBRA TECHNOLOGIES
Slide 12: Future Trends
Future Trends • More and more processing will be at the edge – 5G will help unlock • Complex edge solutions will happen with less processing power e.g. quantization • AI will drive personalization on device e.g. co-pilot • MLOps will mature into a commodity • Winners will be the partners that can dr
Slide 13: Questions
Questions ZEBRA TECHNOLOGIES
Slide 14: Thank You
Thank You ZEBRA and the stylized Zebra head are trademarks of Zebra Technologies Corp., registered in many jurisdictions worldwide. All other trademarks are the property of their respective owners. ©2023 Zebra Technologies Corp. and/or its affiliates. All rights reserved. ZEBRA TECHNOLOGIES