IMU versus INS
An IMU typically combines gyroscopes and accelerometers to measure angular rate and specific force. An INS uses those measurements and initial conditions to estimate position, velocity and attitude. The sensor component and navigation system are not equivalent.
Why does inertial navigation drift?
Bias, noise, scale-factor and initial-attitude errors accumulate through integration. GNSS, vision and odometry can constrain that error. Availability, constraint strength and motion determine the outcome.
Understanding coupling terminology
| Term | Typical information used | Engineering considerations |
|---|---|---|
| Loose coupling | GNSS position and velocity solutions | Clear interfaces; depends on external solution quality |
| Tight coupling | Raw GNSS measurements and inertial data | More detailed use of partial observations |
| Deep coupling | Inertial assistance to receiver tracking loops | Greater receiver and system integration requirements |
“Deep multimodal fusion” is a broader description and does not, by itself, establish receiver tracking-loop integration.
Installation and synchronization
Extrinsic calibration, antenna lever arms, body coordinates and timestamps affect the result. Record orientation, mounting rigidity, vibration and calibration versions, then test actual speeds and turns.
Relationship to HOPO
Pobot Navi combines GNSS/IMU and stereo vision, with sensing options by model. Evaluate continuity, drift, timestamps and recovery together. A continuous update rate is not an accuracy guarantee.