Nebula Trio

AWAREA — Be Aware of Your Surrounding Area

💎 Idea

AWAREA is a water-quality monitoring platform for saltwater bodies, currently covering the Black Sea, but applicable worldwide. The name reflects our mission: make people aware of the water they're about to swim in, fish from, or sail through.

We track both optical parameters (chlorophyll-a, algae blooms, suspended sediment, turbidity) and non-optical parameters (dissolved oxygen, phosphate, nitrate). Non-optical indicators are harder to detect but they determine whether water is healthy or turning into a dead zone — critical for marine life, fishing, tourism, and public health.

The platform serves two modes:

• Beach mode - for swimmers and beach-goers who want to know if the water is safe and pleasant right now
• Offshore mode - for fishing, sailing, and diving, where chemistry and sea-state matter as much as comfort

We pull from Copernicus Marine (water-quality + biogeochemistry + wave/current forecasts), the European Environment Agency (EEA) for the global bathing-water registry, Open-Meteo (weather), live Sofar Spotter buoys operated by IO-BAS and NIMH, and AIS vessel-tracking feeds so users can see what other boats are doing in the area. We fuse these sources into a single 0–100 score per location with a 95 % credible interval, presented through a clean interface.

🛰️ EU space technologies

Data Sources

• Copernicus Marine Service - high-resolution Sentinel-2 ocean colour (chlorophyll/turbidity/SPM at ~100 m) plus biogeochemistry forecasts (dissolved O₂, pH, nitrate, phosphate at 2.5 km)
• Copernicus Marine wave & current forecast - hourly 24-hour forecast of wave height, period, direction, and surface currents at 2.5 km
• EEA Bathing Water Directive registry - worldwide catalogue of coastal and inland bathing waters (~22 000 locations across the EU and partner countries) with names, coordinates, water category, profile links and the EEA quality classification (Excellent / Good / Sufficient / Poor) plus per-season historical microbiological assessments. Each entry carries metadata that's structurally true for that location across seasons — operator, water type, sampling regime, profile documents. Pulled directly from the EEA DiscoMap ArcGIS service that powers the EU's public bathing-water map. We currently surface the ~92 Bulgarian coastal entries, with the rest of the world available on the same endpoint and ready to scale to as we expand.
• Sofar Spotter buoys (IO-BAS / NIMH) - live in-situ water temperature, wave height, wind every 30 min
• AIS vessel-tracking - live positions of commercial ships and fishing boats in the area, surfaced through the platform's vessel layer
• Open-Meteo - weather context

Our Focus

We specialise in fusing modeled (satellite + numerical forecast) and measured (buoy + AIS) sources through a Bayesian time-decay model. Each source is weighted by recency and reliability, with chemistry, sea-state, and live measurements blended into a single defensible number. The credible interval tightens as independent sources agree.

The EEA registry serves as the canonical beach catalogue — globally. It's how we know which beaches exist, where they are officially located, what their legal microbiological classification was last season, and what static metadata (operator, water type, profile documents) describes them year-round. That gives every beach card on the platform a defensible identity and a long-term reference baseline against which our real-time score can be compared. The same source feeds beaches anywhere in the EEA's coverage, so the platform's beach catalogue scales beyond Bulgaria without code changes.

AIS positions enrich the picture with what's happening at sea. Vessels are filtered and weighted by their reported parameters — ship type (tanker, cargo, fishing, passenger, pleasure craft), navigational status, speed over ground, size, and proximity to the location of interest. A drifting fishing trawler near a swimming beach gets surfaced differently than a passing tanker. This lets the offshore view distinguish active fishing zones from transit corridors, and the beach view flag unusual nearby commercial activity.

As we accumulate historical data across seasons, we plan to layer additional capabilities on top of the Bayesian baseline: anomaly detection to flag unusual conditions before they become incidents, and supervised neural-network models trained against ground-truth lab samples (E. coli, in-situ chemistry) and beach-closure events to move from current-state scoring to short-horizon prediction. The current statistical core stays explainable; the learned models will sit alongside it and improve over time as the dataset grows.

Upcoming Enterprise Features

Beyond the consumer-facing modes, we're planning a paid tier aimed at commercial operators:

• Fleet dashboards — multi-vessel monitoring with per-ship water-quality, chemistry, and AIS context for even more context.
• Custom alert rules — webhook + email notifications when chosen indicators cross operator-defined thresholds (bloom forming, hypoxia onset, illegal fishing intrusion)
• Historical analytics & exports — multi-year trend reports, CSV/Parquet downloads, and scheduled report delivery for environmental compliance
• API access — direct REST + WebSocket endpoints for integration into vessel management systems, port authority dashboards, aquaculture farm monitoring, and tourism-board portals
• Per-region SLAs — guaranteed refresh cadence and data availability for paying customers (municipalities, fisheries, hotels, marinas)
• Pollution-event attribution — combining AIS history with chemistry deltas to surface likely sources when a pollution event occurs

How We Use It

For the beach view: the score is a function of chlorophyll, turbidity, suspended matter, water temperature, waves, wind, and forecast sea-state for the next few hours, with the EEA classification and beach metadata surfaced alongside as the official long-term reference.

For the offshore view: chemistry (pH, oxygen, nutrients) layers on top because at km-scale it actually varies meaningfully, and the AIS layer shows nearby vessel traffic filtered by ship type and behaviour.

The work is translating raw satellite parameters and noisy buoy pings into something actually usable — from "chl = 1.7 mg/m³" to "score 78, conditions good, water still cold".

🌊 EU Space for Water

Challenge: Water Pollution Monitoring

We picked this challenge because seas and oceans are where rivers converge. The Black Sea receives runoff from across Eastern Europe, making it a critical monitoring point. What we learn here applies to coastal marine environments worldwide.

How We Contribute

• Dead-zone awareness — tracking dissolved oxygen and nutrient pollution (phosphate, nitrate) in offshore zones surfaces eutrophication risk before major ecosystem damage
• Complete pollution picture — combining optical and non-optical data exposes both visible pollution (algae blooms, sediment) and invisible threats (oxygen depletion, agricultural runoff)
• Public safety — beach-goers get a straightforward score and a "should I go now?" recommendation before getting in the water
• Economic support — tourism and maritime sectors benefit from transparent, accessible data

We make the invisible aspects of water quality visible and actionable for everyone — from weekend beach-goers to operators on the water.

👥 Team

Martin Todorov — Satellite & space data. Works at a company focused on telescope technology and handles our integration with Copernicus and space-based observation systems. Off-hours: skiing and soccer.

Daniel Petrov — Full-stack developer, 10+ years. Leads technical architecture and backend. Has been deep in AI over the past few years and brings that to how we process and present complex datasets. Off-hours: poker & chess.

Ana-Maria Boteva — Product & frontend. The youngest on the team, with a background in economics and a focus on building clean, functional interfaces. Handles design and frontend implementation. Off-hours: trading.

Why We Work Well Together

We've known each other for 5+ years and bring complementary skills across programming, economics, design, and space. A solid range that lets us move quickly from data processing to user interface without big gaps.

Link to our presentation: https://canva.link/ci5un23aojsvob1