System Documentation
System Overview
This system is a mobile bathymetric survey platform that collects depth measurements along a trajectory and reconstructs a terrain surface from filtered spatial samples.
It is designed as a self-contained data acquisition and mapping pipeline, combining real-time sensing, validation, and post-processing.
The system operates in three stages:
- Acquisition — sensor data is collected continuously
- Validation — measurements are evaluated and filtered in real time
- Reconstruction — a surface is generated from selected data points
Data Acquisition
The Arduino-based logger continuously samples:
- GPS — latitude, longitude, and UTC time
- Sonar — depth measurement below the sensor
- Temperature Sensor (DS18B20) — water temperature
- IMU (BNO085) — pitch and roll
Each measurement includes:
- timestamp
- geographic position
- system state
Real-Time Data Filtering
Each measurement is evaluated before being used for mapping.
GPS constraints:
- valid fix
- minimum satellites
- HDOP below threshold
- fix age within limit
Motion constraints:
- speed below threshold
- pitch within limits
- roll within limits
Depth constraints:
- valid reading
- within expected range
- stable vs recent values
Spatial constraint:
- minimum spacing between accepted points
Dual Dataset Design
The system generates two files:
dataXX.csv
- full log
- all measurements
mapXX.csv
- filtered dataset
- mapping-quality points only
Depth Processing
Smoothing:
- moving average filter
- reduces noise and spikes
Sound speed correction:
- based on temperature
- improves depth consistency
Surface Reconstruction
Mapping data consists of:
(lat, lng, depth)
A continuous surface is generated using interpolation.
Inverse Distance Weighting (IDW):
- closer points have more influence
- farther points contribute less
Mapping Resolution
Depends on:
- sampling density
- survey coverage
- platform speed
- filtering thresholds
Coordinates and Units
Coordinate system:
Units:
- depth: cm
- temperature: C
- speed: km/h
Outputs
Raster maps:
- scatter
- interpolated surface
Contour maps:
Geographic outputs:
Error Sources
Sensor:
- beam spread
- noise
- range limits
GPS:
- positional error
- HDOP variation
Motion:
- tilt effects
- speed spacing
Environment:
- temperature variation
- acoustic interference
System Behavior
The system does not generate maps in real time.
It:
- collects filtered samples
- reconstructs surfaces afterward
Design Decisions
Real-time filtering:
- improves data quality early
Dual files:
- preserves raw data
- enables reprocessing
Interpolation:
- handles irregular sampling
- efficient and stable
Limitations
- single-beam sampling
- no salinity correction
- GPS dependency
- interpolation assumptions
Future Work
- pressure depth sensor
- multi-beam sonar
- adaptive sampling
- real-time mapping
- wireless telemetry
- improved interpolation