Underwater acoustic monitoring system LoggLaw NAD-W

Name: LoggLaw NAD-W
Brand: LoggLaw

Underwater acoustic monitoring system with cloud AI analysis. A broadband hydrophone with up to 192 kHz sampling captures underwater sound with high precision and automatically detects marine mammal clicks/whistles and vessel noise (vessel sounds require site-recorded training data). With the optional LTE module, NAD-W gains Edge AI and near real-time monitoring.

The LoggLaw NAD-W set: pressure-housing unit, hydrophone, and antenna — The LoggLaw NAD-W set: pressure-housing unit, broadband hydrophone, and communications antenna

LoggLaw NAD-W overview

LoggLaw NAD-W is a fixed-point underwater acoustic monitoring system that is deployed in the sea, records underwater sound, and uses AI to detect dolphin calls and vessel noise automatically. It was developed by Biologging Solutions Inc. in Kyoto, Japan. Because the original audio is also kept on the SD card, each AI decision can be checked afterward and the AI can be retrained on sound from that site.

Type: A recorder for passive acoustic monitoring (PAM) that stays at one point in the sea and records continuously. It emits no sound of its own; it only listens.
Recording: Broadband hydrophone, 192 kHz / 16-bit / mono (sound up to 96 kHz). Unfiltered WAV files are saved to an industrial SD card (256 GB standard, 512 GB optional).
Detection: AI decides, segment by segment, whether the calls of trained target species (such as Irrawaddy dolphin clicks and whistles) or vessel noise are present. Vessel detection requires training data recorded in the target waters.
Getting results: Either retrieve the SD card and upload the recordings to the cloud, or add the LTE module so the on-device AI sends detections every 15 minutes.
Deployment: Float, stand or anchor. The body is 11 cm in diameter and 40 cm long, about 3 kg in air, rated to 600 m, and runs about 2 weeks on the standard battery.
Main uses: Offshore wind environmental impact assessment (pile-driving and operational noise, marine mammal presence) and rare-species conservation (in operation for the Irrawaddy dolphin conservation project on Songkhla Lake, Thailand).
Name: NAD-W stands for Near-real-time Automated Detection and Warning system.

How it works, specifications, deployment planning, and detection in a 12-page PDF catalogLoggLaw NAD-W Catalog (PDF)

Key Features

Broadband Hydrophone

Captures frequencies up to 96 kHz with broadband sound pickup. Sampled at 192 kHz / 16-bit to precisely record diverse acoustic events including marine mammal echolocation and vessel noise.

Automated Cloud AI Detection

Simply upload SD-card audio to LoggLaw Cloud and receive automated detection of a trained target species' clicks/whistles (e.g., Irrawaddy dolphin) and vessel noise (vessel sounds require site-recorded training data). Outputs include spectrograms, detection timelines, and CSV export.

Edge AI + Near Real-time Transmission (Optional)

With the optional LTE data transmission module, AI classification runs on-device and detection results are auto-transmitted to LoggLaw Cloud every 15 minutes, enabling continuous monitoring without waiting for SD card retrieval.

Deep-Rated Long-Term Deployment

Pressure housing rated to 600 m depth. Approximately 2 weeks of continuous operation on the standard (carry-on compliant) battery, with flexible deployment via floats, stakes, or anchors and easy battery swaps for long-term monitoring.

High-Endurance SD Card Storage

Records uncompressed WAV audio to industrial-grade high-endurance SD cards (up to 256 GB standard / 512 GB optional). Raw, unfiltered data is preserved so recordings can be re-analyzed as detection models improve.

Dashboard & Alerts

Visualize detection events on map, timeline, and statistical dashboards via LoggLaw Cloud. Email/SMS alerts for specific events and centralized management of multiple devices.

Your recordings train the AI

Besides the detection results, NAD-W saves the unfiltered audio to the SD card as WAV files, so you can verify detections against the original recordings afterward. The AI is retrained on raw data recorded at your site, and the updated model is deployed to both the Cloud AI and the on-device Edge AI. Detection accuracy improves as the model adapts to the local soundscape.

The improvement cycle

Record
AI detects & alerts
Verify against raw audio
Retrain
Update the modelBack to recording, and repeat

Two ways to use it

A Basic

SD card retrieval → Cloud AI

1Record
2Retrieve SD card
3Upload WAV files
4View results

B With LTE module

Edge AI → near real time via LTE

1Record
2Edge AI analysis
3LTE upload
4View on the map

In both modes, raw audio is saved to the SD card, so after retrieval you can verify detections against the original recordings.

Use in Offshore Wind Environmental Impact Assessment (EIA)

Offshore wind EIA requires continuous assessment and monitoring, from baseline through operation, of how pile-driving and operational noise affect marine mammals. NAD-W and the citizen-science monitoring portal support the full chain from underwater-sound recording to data operations and reporting.

Construction & operational noise monitoring

Broadband recording captures pile-driving, vessel and operational noise over the long term. With the portal's construction-schedule overlay and predicted noise levels, it supports mitigation such as soft-start and work-window coordination.

Passive acoustic monitoring (PAM)

Automatically detect dolphin/whale vocalizations and vessel traffic (vessel sounds require site-recorded training data). The AI can be extended to additional species through training, supporting presence/absence and spatio-temporal distribution analysis.

Long-term, spatial monitoring & control-site comparison

Deploy multiple NAD-W buoys across both impact and control areas, with voltage, detections, maintenance history and signal coverage managed in one place. By comparing the impact–control difference across baseline / construction / operation phases, it supports study designs (BACI) that isolate construction-driven change.

Evidence & reporting

Analyze KDE hotspots, in/out-of-MPA ratios and environmental correlations. Export sighting, detection and survey data as CSV to support assessment reports and post-construction monitoring.

Configuration for target species & regulations

NAD-W records up to 96 kHz (192 kHz sampling). In waters where high-frequency echolocators such as harbor or finless porpoise (~125–140 kHz) are the primary receptors, we recommend pairing NAD-W with a dedicated high-sample-rate PAM, with NAD-W covering broadband noise, vocalizing species and the integration platform. For regulatory reporting of absolute sound levels (dB re 1 µPa), calibrated measurements are used. We propose the optimal configuration for your target species and site.

Discuss offshore-wind monitoring (joint research / contract)

About our commissioned marine acoustic monitoring and joint research

Specifications

Detailed Technical Specifications

Hardware Specifications

Dimensions	Diameter 11 cm × Length 40 cm (main unit)
Weight	Approx. 3 kg in air (with integrated battery)
Depth Rating	600 m
Operating Temperature	-10°C to +50°C
Battery Life	Approx. 2 weeks (standard battery, carry-on compliant)
Memory Capacity	256 GB (standard) / 512 GB (optional)
Recording Duration (Continuous)	256 GB: ~185 hours (~7.7 days) 512 GB: ~370 hours (~15.4 days) * at 192 kHz / 16-bit / Mono
Recording Duration (50% Duty)	256 GB: ~15.4 days 512 GB: ~30.8 days
Data Interface	Removable industrial-grade SD card. System configuration via config file written to the SD card. No proprietary reader or cable connection required.
Connectors	Underwater connectors for hydrophone, LTE antenna, and GPS antenna (easy attach/detach)
Deployment	Compatible with floats, stakes, and anchors; easy battery replacement.

Recording Specifications

Sampling Rate	192 kHz
Frequency Range	Up to 96 kHz (Nyquist frequency)
Audio Resolution	16-bit
Channels	Mono
Data Format	WAV with embedded timestamps
Recording Mode	Continuous, duty-cycle, and scheduled recording are supported. With the LTE module, an alternating 30 s recording / 30 s analysis cycle runs the Edge AI on the device.
Signal Filtering	Raw audio is recorded without filtering. All detection and classification is performed by the AI engine, preserving original data for re-analysis with improved models.

AI Capability

Detection Targets (extensible via training)	Detects presence/absence of a trained target species' vocalizations per recording segment. Irrawaddy dolphin clicks/whistles and vessel noise are examples; train an acoustic model per target to extend the classifiable set (vessel sounds require site-recorded training data). Vessels can also be classified by type (long-tail boat, cargo vessel, etc.).
Custom Models	Additional detection models for dugong vocalizations, snapping shrimp, fish choruses, and more (discussed separately)

LTE Data Transmission Module (Optional)

Edge AI Analysis	Real-time on-device acoustic classification (Irrawaddy dolphin, vessel noise, etc.; vessel sounds require site-recorded training data)
Communication	3G/4G LTE (MQTT protocol). SIM card and cellular data plan to be procured by the customer.
Data Transmission Interval	Every 15 minutes
GPS	Daily position fix (supports tracking of dozens of devices)
Behavior During Communication Loss	On-device data caching during communication interruptions; automatic retransmission upon recovery

Recording times are based on SD card capacity; a deployment ends when the battery runs down or the card fills, whichever comes first (e.g., continuous recording fills a 256 GB card in about 7.7 days, before the battery runs down).

Specifications are subject to change without notice.

* Specifications are subject to change without notice for improvement purposes.

Connected Citizen-Science Monitoring Portal

Acoustic data collected by NAD-W can be aggregated into a cloud-based citizen-science monitoring portal. Sightings, stranding reports, device status and analysis are unified on a single map that connects researchers, authorities and citizens on one operational board. It can be configured for any target animal, from terrestrial GPS collars to marine acoustic buoys, and the Irrawaddy dolphin conservation project on Songkhla Lake, Thailand is one such example.

What the portal does

Live monitoring map

Layer sightings, vessel/construction noise, protected areas and sensor status to grasp the whole site in near real time (citizen sighting reports are immediate; device data arrives via LTE every ~15 min, or after SD-card retrieval without LTE). Timeline playback reconstructs past detections chronologically.

Acoustic device monitoring

Manage voltage, daily detection counts, last-seen and status for every NAD-W buoy in one place. Low-battery and offline conditions trigger browser-push and LINE alerts, with thresholds set per project.

Citizen sighting & stranding reports

Report sightings or strandings from mobile with one-tap GPS or a map pin. Multilingual UI, photo attachment and anonymous posting are supported, and emergencies are routed to response teams instantly.

Offline-first field surveys

Record location, weather and observation events on-device even without connectivity at sea or in the field. Data auto-uploads on reconnection and behaves identically on iPhone and Android.

Statistics & deep analysis

Beyond KPIs like total sightings, group size and seasonal trends, analyze KDE hotspots, in/out-of-MPA ratios and acoustic breakdowns by target species / vessel / construction noise.

Notifications & data integration

Push key events via browser and LINE. Export sighting, detection and survey data as CSV, and securely share read-only feeds with partner organizations.

Portal screens

Live monitoring map: sightings, sensors and protected areas on one map

Acoustic device monitoring: NAD-W buoy status and detection counts at a glance

Statistics dashboard: sightings, individuals and seasonal trends

Deep data analysis: population, spatial, behavior, acoustic, environment, device

* Screens show the platform running for the Irrawaddy dolphin conservation project on Songkhla Lake, Thailand. The same foundation can be configured for other target animals such as monkeys or deer.

Explore the citizen-science portal

Start small, validate together

Underwater sound differs from site to site, so the AI performs best once it has been tuned to your site's soundscape. We start with a few units to collect field recordings, refine the detection targets and accuracy with you, and then scale up.

STEP 1
Record on site
Deploy a few units and record in the target area.
STEP 2
Tune the AI
Review the raw data together to refine targets and improve accuracy.
STEP 3
Scale up
Once the results are confirmed, add units and expand coverage.

Interested in other species or sounds? Contact us; we can tune the AI during the pilot.

Frequently asked questions

What is LoggLaw NAD-W?

A fixed-point underwater acoustic monitoring system deployed in the sea to record underwater sound. It records with a broadband hydrophone at 192 kHz / 16-bit, and cloud AI automatically detects the calls of trained target species (such as Irrawaddy dolphin clicks and whistles) and vessel noise.

Up to what frequency can it record?

With 192 kHz sampling it records sound up to 96 kHz. Where the main targets are high-frequency echolocators around 125–140 kHz, such as harbour porpoises and finless porpoises, we recommend pairing it with a dedicated high-sample-rate PAM recorder, with NAD-W covering broadband noise and vocal species.

How long can it monitor?

About two weeks of continuous operation on the standard battery. Continuous recording at 192 kHz / 16-bit / mono fills 256 GB in about 185 hours (about 7.7 days) and 512 GB in about 370 hours (about 15.4 days). At a 50% duty cycle, 256 GB holds about 15.4 days and 512 GB about 30.8 days. A deployment ends when the battery runs down or the card fills, whichever comes first; with continuous recording to 256 GB, the card fills first. The battery is easy to replace.

What sounds can it detect automatically?

It decides, segment by segment, whether the calls of trained target species are present. Examples are Irrawaddy dolphin clicks and whistles and vessel noise; vessels can also be classified by type, such as longtail boats and cargo ships. Detecting vessel noise requires training data recorded in the target waters. Additional models, such as dugong calls, snapping shrimp and fish choruses, are available on request.

Does detection performance depend on the site?

Yes. AI detection performance depends on the acoustic environment at the site, such as background noise and the distance to the animals. That is why the AI is retrained on raw data recorded at your site and the updated model is deployed, so accuracy improves as the model adapts to the local soundscape.

Can we see results without recovering the unit?

With the optional LTE data module, edge AI analyzes sound on the unit and sends detections to the cloud every 15 minutes. You provide the SIM card and data plan. If the connection drops, data is cached on the unit and resent automatically once it recovers. Without LTE, recordings on the SD card are uploaded to the cloud for analysis.

How is it deployed?

It supports float, stand and anchor deployments. The body is 11 cm in diameter and 40 cm long, weighs about 3 kg in air including the battery, and is rated to 600 m. Settings are made in a configuration file on the SD card, so no dedicated reader or cable is needed.

Can it be used for offshore wind environmental impact assessment?

Yes. It records piling, vessel and operational noise over long periods and automatically detects dolphin and baleen-whale calls. Multiple units can be placed in impact and control areas to compare before, during and after construction (BACI design). When regulations require absolute sound pressure levels (dB re 1 µPa), calibrated measurement is used.

Can we start with just a few units?

Yes. We start with a few units to collect field recordings in the target area, and review the raw data with you to refine the detection targets and accuracy. Once the results are confirmed, we recommend adding units and expanding coverage.

Can we commission the monitoring and analysis itself?

Yes. Apart from buying the equipment, you can engage us for commissioned marine acoustic monitoring or joint research. See the marine acoustic monitoring page for details.

How does it differ from LoggLaw CAM?

NAD-W is a fixed-point device: it is deployed at one spot in the sea and keeps recording underwater sound there for a long period. LoggLaw CAM is an underwater video logger, mainly attached to animals, that records video and hydrophone audio together. NAD-W suits continuous monitoring of one area; CAM suits studying an animal's behavior together with the sounds around it.

What does the name NAD-W stand for?

Near-real-time Automated Detection and Warning system. The name describes what the system does: it records underwater sound, detects target sounds automatically, and sends notifications when needed.

Inquire About LoggLaw NAD-W

For detailed product information, pricing, case studies, and more, feel free to contact us. Our expert staff will propose the best solution for your needs.