While standard procedures ensuring accurate multi-sensor integration within a multibeam system already exist, operators are still plagued by small periodic systematic residuals in data. The source of these errors has long been recognized as the result of imperfect definitions of offsets, alignments, latencies and sound speed between the sonar and auxiliary sensors. In isolation, and shallow water geometry, any one of these sources is easy to recognize. As either water depth or the number of error sources increase, however, the resulting pattern of residuals can confound simple analysis. Gross errors are usually immediately apparent, but small errors such as a few milliseconds of latency or decimeters of lever arm, particularly in combination, can be hard to identify.

As a continuation of the Rigorous Inter-Sensor Calibrator (RISC), a theoretical model-based method previously developed, the approach has now been refined and applied to real-time data. The method relies on separating the bathymetric residuals in the seafloor data from the underlying truth in order to parametrically estimate the mismatch. Using a georeferencing model that explicitly includes the impact of six possible integration errors, the difference between the integrated surface, and a smoothed version of it, is minimized. Initial results have shown sensitivity of the method to variations in seafloor bathymetry and external systematic errors such as imperfect real time heave and water column modeling. To mitigate these issues more robust surface modelling and optimization techniques are proposed. To facilitate the real-time implementation of this tool, an improved multibeam integration model is also proposed.