SwarmMap - Scaling Up Real-time Collaborative Visual SLAM at the Edge
Offered By: USENIX via YouTube
Course Description
Overview
Explore a cutting-edge framework for scaling up real-time collaborative visual SLAM at the edge in this 17-minute conference talk from NSDI '22. Dive into the design and implementation of SwarmMap, a system that addresses the challenges of data explosion in multi-agent scenarios. Learn about the three core modules: a change log-based server-client synchronization mechanism, a priority-aware task scheduler, and a lean global map representation. Discover how SwarmMap enhances performance and capacity in edge-based multi-agent visual SLAM applications, serving twice as many agents as existing solutions while maintaining superior accuracy. Gain insights from comprehensive evaluations and a real-world case study at a major oil field, demonstrating SwarmMap's practical benefits for emerging mobile applications in search-and-rescue, inventory automation, and drone grouping.
Syllabus
Intro
Background
Scalability issues
Key Insight
Mapit - Map Information Tracker
MBP - Map Backbone Profiling
STS - SLAM-Specific Task Scheduling
Experiments
Conclusion
Taught by
USENIX
Related Courses
Scaling Memcache at FacebookUSENIX via YouTube Multi-Person Localization via RF Body Reflections
USENIX via YouTube Opaque - An Oblivious and Encrypted Distributed Analytics Platform
USENIX via YouTube Live Video Analytics at Scale with Approximation and Delay-Tolerance
USENIX via YouTube Clipper - A Low-Latency Online Prediction Serving System
USENIX via YouTube