Robots Atlas>ROBOTS ATLAS
Infrastructure

Round-trip latency

1983ActivePublished: 17 November 2026Updated: 17 November 2026Published
Key innovation
Framing link responsiveness as the full request-response loop, making RTT a direct determinant of the safety and smoothness of real-time controlled systems.
Category
Infrastructure
Abstraction level
Pojęcie
Operation level
ServingDeployment
Use cases
Telesurgery and robot teleoperationNetwork diagnostics (ping, traceroute)Tuning TCP parameters and the send windowOnline gaming and real-time systemsRemote control of vehicles and machines

How it works

The sender transmits a packet (or control stimulus) and measures the time until an acknowledgment or response arrives. The total comprises round-trip propagation (physically bounded by the signal speed in the medium - about two-thirds the speed of light in fiber), transmission time, queuing and remote-side processing. Protocols such as TCP estimate RTT (a weighted average; the Jacobson/Karels algorithm also accounts for standard deviation) to tune retransmissions and the send window. A large bandwidth-delay product requires mechanisms such as TCP window scaling.

Problem solved

Interactive and control systems (teleoperation, robot control, network protocols) need a measure of the link's real responsiveness in both directions, not just one way. RTT captures the full loop: request and response.

Evolution

1983
RTT in TCP
Inflection point

TCP standardization made RTT estimation foundational to retransmission and flow control on the Internet.

1988
Jacobson/Karels algorithm

Improved RTT estimation accounting for standard deviation enhanced congestion-control stability.

2001
RTT as the telesurgery barrier

The Lindbergh Operation showed that low, stable RTT - not bandwidth alone - decides the feasibility of remote surgery.