Key finding
The study catalogued 255 digital assets, defined as automated monitoring products that collect or process biodiversity data without requiring a human in the loop at the point of collection. Satellite-derived assets dominate at 68 percent, but 85 percent of assets monitor only primary producers (plants, phytoplankton, algae), and 82 percent resolve only to phylum or kingdom level, not species. Only 15 percent provide species-level data (38 assets total; 28 for animals). Arthropods, the vast majority of animal species, have just 13 percent species coverage.
North America and Europe have the highest concentration of assets, while the most biodiverse countries in Africa and Asia have the fewest. Over half of assets with daily resolution have data latency exceeding one month, and 70 percent of annual-resolution assets have latency over one year.
What the research asks
Automated biodiversity monitoring, using satellites, acoustic sensors, camera traps, and environmental DNA, has been hailed as a transformative solution for tracking the global biodiversity crisis. Yet despite years of technological advancement and falling hardware costs, automated monitoring remains far from operational at the scale needed. This study asked why the promise of automated biodiversity monitoring has not translated into widespread use in policy and management, and what the specific bottlenecks are, so that researchers, funders, and technology developers can direct their efforts where they will have the greatest impact.
What it finds
The study identified 255 digital assets, defined as automated monitoring products that collect or process biodiversity data without requiring a human in the loop at the point of collection. The results reveal a dramatic mismatch between where monitoring capacity exists and where it is most needed.
Satellite-derived assets dominate, accounting for 68% of all monitoring products. But 85% of these assets monitor only primary producers (plants, phytoplankton, algae), and 82% can identify organisms only to the level of phylum or kingdom, not species. Only 15% of assets (38 in total) provide species-level data, and for animals the number drops to 28. Arthropods, which make up the vast majority of animal species on Earth, have just 13% species coverage.
The spatial picture is equally stark. North America and Europe have the highest concentration of monitoring assets, while countries in Africa and Asia with the highest threatened species richness have the fewest. In the marine realm, assets are concentrated around North America, Europe, and Oceania, leaving the Tropical Atlantic, Western Indo-Pacific, Southern Ocean, and Arctic with the largest monitoring gaps.
Temporal latency is another major bottleneck. More than half of all assets with daily or higher temporal resolution have a data latency of more than one month, meaning the data arrives too late for real-time decision-making. Seventy percent of assets with annual or multi-year resolution have a latency exceeding one year. Species-level assets fare somewhat better on this measure, suggesting that when species-specific data is collected, it tends to be processed more urgently.
The authors propose a four-step framework for addressing these gaps: developing new sensor technology for underrepresented taxa, deploying sensors more strategically in biodiversity-rich regions, investing in machine learning and AI for data processing, and creating robust data sharing infrastructure.
Why it matters
This paper delivers a sobering reality check alongside a practical roadmap. Companies and governments making biodiversity commitments under frameworks like the TNFD (Taskforce on Nature-related Financial Disclosures) and the Global Biodiversity Framework need reliable monitoring data to report progress. But this study shows that the automated tools they might rely on cover only a fraction of biodiversity and are heavily biased toward wealthy countries.
The four-step framework offers clear investment priorities for anyone building biodiversity monitoring systems, from conservation NGOs to corporate sustainability teams. The finding that data processing and sharing bottlenecks may be the easiest and cheapest to fix, requiring less effort than building new sensors, is a call to action for the open data and technology community. For Earth PBC's audience, this paper underscores the importance of integrating diverse data sources into a unified monitoring platform: satellite observations for vegetation, acoustic sensors for wildlife, and community-collected data for ground truth.