Automation Driven Marine Geophysical Data Processing Center

The use of automated SSS object detection reduced the time required to identify objects by approximately six times, significantly increasing processing efficiency across a large-scale offshore survey.

This was our first survey conducted in our hometown. As part of the dredging and desilting works on the Bosut River, an object detection survey was undertaken, resulting in the identification of 694 man-made objects. The survey contributed to supporting the river restoration works and promoting a cleaner, healthier aquatic environment.

The survey was conducted in shallow waters (up to 25 m) characterised by a mobile seabed shaped by strong tidal currents. Sand waves are present along the route, accompanied by notable variations in acoustic reflectivity. In the nearshore section, pUXO targets were identified, indicating a high probability of WWII legacy ordnance. SBP data also indicates the potential presence of shallow gas.

As part of a project for the Ministry of Maritime Affairs, Transport and Infrastructure, a cross-sectional survey of the Sava River was conducted to assess riverbed conditions and calculate the volume of material available for extraction.

Variations in the physical properties of the water column resulted in two datasets of significantly different quality. The section closer to Bornholm Island exhibited minimal noise, whereas the section closer to the Danish coast was heavily affected by thermocline. These conditions required the development and implementation of a new data processing workflow.

Data processing wasparticularly challenging due to the environmental conditions.

The SSS data exhibited positional offsets of up to 100m relative to the MBES data, while the MBES data showed significantsediment mobility. Despite these challenges, the required detection accuracyand client acceptance criteria were successfully achieved.

The survey was conducted in shallow waters (up to 45 m water depth) over a generally flat and stable seabed. The nearshore section is characterised by variations in acoustic reflectivity, changes in sediment type, and numerous man-made features associated with intensive human activity. Potential UXO (pUXO) targets were identified, indicating a high probability of WWII legacy ordnance.

Located within the Niger River Delta, the survey area was interpreted as a stable environment with a predominantly flat seabed. As shallow gas is a recognised regional geohazard within the Niger Delta, was therefore specifically considered during interpretation of the SBP data. No targets, shallow gas indicators, or other geohazards were identified; however, the presence of diffuse, low-saturation gas cannot be completely excluded.

The project involved surveying the dredging area and re-routing the power cable to support the construction of an artificial island in the Arabian Gulf. Geological interpretation was constrained by ghost reflections and limited sub-bottom penetration caused by the sediment characteristics; however, the interpreted horizons were successfully correlated with the available geotechnical data.

The geophysical survey data revealed a high density of boulders within the survey corridor. Target-picking criteria was stringent regarding targets position, so all detected targets were shifted to the corresponding MBES location. Among the identified targets, one non-dangerous wreck was detected, although it was not visually conspicuous.

The survey covered shallow waters (up to 25 m) with a highly mobile seabed shaped by strong tidal currents. Operations were complicated by very shallow sections at the Reine Delta entrance, potential UXO hazards, and heavy marine traffic.

The geophysical survey data revealed a high density of boulders within the survey corridor. Boulder fields were classified, and the target-picking criteria were adjusted to meet the client's requirements. Among the identified targets, one non-dangerous wreck was detected, although it was not visually conspicuous.

‍

This project marked the first application of our automated object detection solution, originally developed for Side Scan Sonar (SSS), to Synthetic Aperture Sonar (SAS) data.

With only minimal modifications, the solution was successfully applied to the significantly more detailed and complex SAS datasets, demonstrating the robustness and flexibility of the methodology.

Several wrecks were detected, fishing ship, steam ship, military destroyer (HMS LYNX sunk in 1915).

The survey was conducted in shallow waters (up to 45 m depth) characterised by a flat and stable seabed. Nearshore part is characterized with boulder fields with notable variations in acoustic reflectivity which can be a challenging factor for cable lay. The pUXO targets were identified, indicating a high probability of WWII legacy ordnance.

Pre-lay geophysical route survey within the German 12 nm zone, parallel to the coastline. The route traverses a dynamic seabed characterised by extensive sand waves and other morphological features, with numerous existing subsea assets, presenting significant challenges for cable installation.

Surveys in shallow waters (up to 20 m) with mobile seabed conditions, area shaped by strong tidal currents. Survey on a shipping route into Ems estuary, west of Borkum island, proximity to lot of wrecks. The waters west and northwest of Borkum are considered a high-probability WWII legacy area.

The project comprises the Princess Elisabeth Energy Island, six export cables to the mainland, and surveys in shallow waters (up to 35 m) with mobile seabed conditions. UXO risk maps identify areas of potential 25 kg and 50 kg ferrous masses.

The survey area comprised shallow water with a generally flat seabed but exhibited distinct heterogeneous seabed conditions. Although the surrounding area is dominated by sand deserts, the eastern section is characterised by thin unconsolidated sediment overlying outcropping bedrock, which was not expected.

This project marked the first large-scale implementation of automated SSS target rationalization. More than 2.2 million individual object detections were identified across overlapping SSS survey lines. Using an automated rationalization process, processors correlated repeated detections of the same object and calculated a single averaged position and dimensions for each target, maximizing the precision of the final target list.

To meet the client's strict positioning requirements, an automated workflow detected all targets in the SSS data and shifted their positions to the corresponding MBES locations, ensuring that every detected target was assigned the most accurate available position.