About
The vast majority of hydrographic surveys are performed using a GNSS-aided inertial navigation system (GNSS/INS) which provides precise time, latitude, longitude, height, pitch, roll, and yaw of the vessel and sensor(s). But what if we need to perform a hydrographic survey in a GNSS-denied environment? Historically the best solution was to rely on robotic total stations, but that approach has significant limitations with respect to line-of-sight and time synchronization. Point cloud generation using SLAM technology is becoming increasingly common. While some efforts have been made to use SLAM methods with acoustic sensors such as multibeam echosounders, there are limitations involved here as well. Acoustic data tends to have a much lower signal-to-noise ratio than LiDAR or Imaging sensors along with a wider beam width and therefore lower resolution. Further, most multibeam echosounders are configured in a 2D array (swath) while SLAM methods generally rely on rotating LiDAR sensors or 360° camera arrays. Instead of using SLAM directly with the acoustic data, our solution was to use a SLAM LiDAR system on the survey vessel as a navigation device in lieu of GNSS. While the most visible result of a successful SLAM survey is a cohesive point cloud, the relative position and orientation of the LiDAR sensor is also calculated. By extracting and formatting this position and orientation data to simulate a post-processed trajectory file (i.e. SBET) it can be applied to multibeam or any other sensor on the vessel in an otherwise standard workflow. I will discuss the original concepts and in-house testing we performed to bring this idea to reality and highlight a survey performed at Pier 39 in San Francisco which required high-accuracy multibeam data to be collected underneath an existing pier. We will share the methodology, the results of the survey, and the lessons learned. We’ll also discuss the accuracy and repeatability of the data, the time synchronization considerations, and the applicability of this method to different GNSS-denied environments.

