Future work

Scalability
• More agents per machine
  • Multi-threaded agent server
  • “Hot” interpreters
  • Faster interpreters
• More agents in the network
  • Compression
  • Code caching
• More machines per cloud
  • New versions of APRL
  • New routing algorithms

Active networking
• Lower-level implementation
• Different routing algorithms

Hypothesis tracker
Automatically determine which

observations correspond to the
same vehicle

Previous slide

Next slide

Back to first slide

View graphic version