Aqua guard

The problem. Algae blooms and bacterial contamination contaminate European inland and coastal waters every summer. In 2024 alone, Berlin alone closed bathing sites at Müggelsee nine times. By the time a water sample reaches the lab — typically 48 hours later — bathers are already exposed to cyanotoxins that exceed the WHO microcystin limit by up to 700x in German waters. EU economic loss: €4.4 billion per year across tourism, fisheries, drinking water treatment, and public health. Authorities react after the bloom hits, because they can't see it coming.

Our solution: AquaGuard — a closed-loop monitoring system. Software continuously watches Copernicus satellite feeds for the customer's water bodies. The moment the AI flags a chlorophyll anomaly, standing-by autonomous drones receive a Galileo-guided mission, fly to the hotspot, trace the bloom to its source with onboard RGB and thermal cameras, and deploy a floating live-analysis pod that streams chlorophyll, temperature, and dissolved oxygen in real time. That fresh ground-truth data fuses with the satellite history and weather record, and our per-customer AI forecasts the next bloom event seven days ahead — algae and bacterial risk, mapped to specific hotspots. Every mission sharpens the model. The loop closes. Reactive becomes predictive. A 48-hour response becomes a 90-minute one. We've already trained a working forecast model on three years of Müggelsee data, integrated six Copernicus services with Galileo positioning, and demonstrated autonomous mission execution. Production accuracy reaches 80%+ within 12 months of customer deployment as drone data feeds back into the AI.

🛰️ EU space technologies

AquaGuard is built structurally on Copernicus and Galileo — not as decoration. We currently fuse:

• Sentinel-3 OLCI — chlorophyll-a concentration at 300m, our primary bloom-trigger signal.
• Sentinel-3 SLSTR — surface water temperature for thermal anomaly detection and stratification analysis.
• Sentinel-2 MSI — high-resolution NDCI (chlorophyll proxy) and NDWI (water masking) for pixel-level hotspot detection.
• Sentinel-1 SAR — all-weather backscatter, our cloud-day fallback when optical sensors are blocked.
• Copernicus Climate Data Store (ERA5-Land) — temperature, precipitation, wind, solar radiation, and humidity history feeding our prediction AI.
• CORINE Land Cover — upstream agricultural and urban land use mapped to bloom hotspots, identifying nutrient-loading sources.
• Galileo positioning — EU-sovereign GNSS for autonomous drone navigation, with OSNMA-ready architecture for high-trust deployments.

The value: each data stream contributes a different physical signal, and the fusion is what makes prediction possible. A single satellite gives a snapshot; six fused streams plus weather context give the model enough signal to forecast a bloom seven days before Sentinel-2 confirms it visually. Galileo gives us the operational sovereignty to fly autonomously over critical EU water infrastructure without depending on non-European positioning. Together, this turns a fragmented set of public data products into a single actionable answer for water authorities: "this water body will bloom on these dates, in these locations, at this severity."

🌊 EU Space for Water

We're addressing Challenge #2 — Pollution Tracking and Source Identification.

Current European water-quality response is reactive. Manual sampling teams arrive after a bloom is reported, take samples, wait 48 hours for lab results, then issue closures or treatment orders — often after public exposure has already occurred. AquaGuard contributes to protecting and managing EU water resources in three concrete ways:

1. Predictive prevention. Our AI forecasts both algae blooms and bacterial contamination risk seven days ahead, giving water authorities advance notice to plan treatment, advisories, or closures before public exposure happens. Instead of reacting to a confirmed bloom, authorities prepare for a predicted one.

2. Pollution-source identification. When a hotspot is detected, our drones fly upstream and use onboard computer vision (RGB + thermal fusion) to trace concentration gradients back toward the source — agricultural runoff, urban discharge, sewer overflow. Water authorities receive a geo-tagged report with not just what and where, but why and from where.

3. Continuous ground-truth feedback. Every drone mission and every floating sensor pod feeds new high-resolution data back into the model — sharpening forecasts, building a per-water-body digital twin, and creating an ever-improving baseline for future EU water-quality directives.

This shifts European water management from reactive testing to predictive protection — exactly what Challenge #2 asks for, scaled across the more than 22,000 designated bathing waters and the 127,000+ surface water bodies monitored under the EU Water Framework Directive.

🤼 Team

Skymeteors — Berlin-based autonomous-systems team. 5 members.

• Skymeteors — Berlin-based autonomous-systems team. 5 members.
  • Jack Safia — Founder & technical lead. Mechatronics engineer, Berliner Hochschule für Technik. German-Syrian. Built and flew the autonomous quadcopter platform that AquaGuard runs on. Leads system architecture, mission planning, and EU integration.
  • Leo Safia — Mechatronics engineer. Berliner Hochschule für Technik. Hardware integration, drone airframe, sensor wiring, and field-readiness. Co-built the autonomous platform and the pod-deployment mechanism.
  • M. Hajri — Software developer. Oulu University, Finland. Builds the satellite ingestion pipeline, the XGBoost prediction model, and the operator-facing dashboard. Trained the per-customer AI on Müggelsee historical data.
  • A. Ismail — Software engineer. University of Duisburg-Essen, Germany. helped developing the Jetson AGX Orin computer-vision stack — RGB + thermal fusion, source-direction tracking, ROS Noetic integration.
  • Paul Petras — Business & operations. Business Administration, University of Potsdam. Leads outreach to water authorities, EU funding pipeline, partnership development, and go-to-market strategy.

We're growing: actively seeking an Umwelttechnik / environmental engineering graduate to join the core team, plus advisors from Austria's hydrology, water-utility, and EU water-policy community. If you build, study, or regulate Europe's water bodies — we want to talk.