Robust and flexible robotics messaging middleware

Determine a design and implementation for a robust and flexible robotics messaging middleware that scales with increasing numbers of components, layers, sensors, processors, and network types, and that remains reliable under non-trivial communication delays, high-bandwidth data, wireless congestion and variability, and modern memory hierarchy performance penalties.

Background

Robotics systems increasingly rely on middleware for messaging among distributed components, yet growth in system complexity and resource demands has undermined traditional assumptions such as communication being "fast enough to be reliable." Modern sensors produce orders of magnitude more data, wireless networks exhibit congestion and variability, and memory systems introduce significant performance penalties when traversing complex hierarchies.

These trends make integration more difficult and slow development, highlighting the need for a messaging system that can reliably manage resources and performance across heterogeneous hardware and network conditions. The paper explicitly notes that achieving such a system remains an open question.