The present subject matter discloses a method (700) for estimating a pose of a first sensor at a measurement timestamp of a second sensor operating asynchronously with the first sensor. The method (700) includes receiving a plurality of inputs comprising a plurality of timestamped poses of the first sensor within a predetermined time window, one or more motion-state measurements from a third sensor operating at a higher sampling rate than the first sensor, and a query timestamp corresponding to a measurement of a second sensor. The method (700) includes determining a baseline pose of the first sensor at the query timestamp by computing a closed-form interpolation between the plurality of timestamped poses of the first sensor. The method (700) includes generating a pose residual relative to the baseline pose, an uncertainty estimate associated with the pose residual, and an estimate of timestamp-noise level based on the plurality of inputs. The method (700) includes determining a synchronization-quality gating value as a continuous monotonic function of the timestamp-noise level, to be applied to scale the pose residual to produce a scaled pose residual. The method (700) includes combining the scaled pose residual with the baseline pose to generate a predicted pose corresponding to the query timestamp. The method (700) includes outputting the predicted pose, the uncertainty estimate, and the estimate of the timestamp-noise level. (FIG. 7)