From Spectral Data to Measurement Infrastructure: Scaling Science-Grade Earth Observation

EarthDaily is a global Earth observation company focused on delivering science-grade data and analytics designed for broad-area change detection and decision-centric intelligence.¹ With the EarthDaily Constellation, the company is building a system for daily, globally consistent measurement to support governments and enterprises operating in complex, high-impact environments.¹

Following the release of its first processed imagery from the EDC-01 satellite In February, EarthDaily is moving from early validation into operational scale.² With six additional satellites scheduled for launch in early May, the company is beginning to build out the constellation toward continuous, analysis-ready data delivery, with commercial operations expected later this summer.

In this Q&A, CEO Don Osborne discusses the design of the EarthDaily Constellation, including its 16-imager-per-satellite architecture, calibration-driven approach, and the broader shift from imagery to reliable measurement at global scale.

Can you talk about EarthDaily’s upcoming launches and what the next phase of the constellation looks like?

We are moving from validation to scale. EDC-01 was launched early to prove the system end to end. That includes sensor performance, calibration, and the full data pipeline across acquisition, downlink, and processing. This step matters because once you scale, you are scaling everything. If the foundation is not right, the errors compound.

With six additional satellites launching in May, we begin building real capacity. That means meaningful gains in coverage, revisit, and consistency across geographies. It is the point where the system starts behaving like a constellation rather than a single asset.

Each satellite adds not just more data, but more consistency. That consistency is what enables monitoring instead of simple observation. With daily acquisition at a consistent geometry and time of day, the system is designed to support a broad spectral range at global scale, allowing different conditions and regions to be observed in a coordinated way.

This next phase is the system becoming operational. We are expecting to reach this goal late Summer of 2026.

What drove the decision to use 16 imaging assemblies instead of a single hyperspectral system?

This starts with a simple reality. Customers do not need more imagery; they need reliable measurement. Across agriculture, resource management, and government, the gap was clear. There was no system delivering consistent, daily, wide-area measurement that could be trusted for real decisions. You cannot achieve daily global coverage, high spatial resolution, and measurement stability from a single optical system. Those requirements are fundamentally in tension.

So, we designed the system to solve for all three. Sixteen coordinated imagers allow us to maintain controlled geometry and strong signal quality while achieving a wide swath. That wide swath enables daily global coverage at a consistent local time. The 22 spectral bands were selected not just for richness, but to support atmospheric correction, cloud and haze detection, and consistency across varying conditions. That is what allows measurements to remain comparable over time.

The trade-off is complexity. Sixteen optical paths introduce variability that must be continuously managed, so the system behaves as a single instrument. That complexity is intentional. It is what allows us to deliver consistent measurement at planetary scale.

How does EDC-01 operate as a coordinated measurement platform?

At a hardware level, there are sixteen instruments. At a system level, it must behave like one. Each imager has its own optical path, detector characteristics, and thermal profile. As the satellite moves through orbit, conditions are constantly changing. Without control, those differences would introduce noise and inconsistency. The work is in removing that variability. It begins at acquisition with coordinated timing, pointing, and coverage planning to ensure consistent observation geometry.

The more complex work happens in calibration and processing. Every source of variation, including detector response, thermal drift, and atmospheric effects, is measured and corrected continuously. What makes this possible is system integration. Satellite operations, ground systems, and processing pipelines are designed as a single architecture. At scale, with thousands of acquisitions per day, that coordination must be automated.

The outcome is simple. A measurement taken today is directly comparable to one taken tomorrow, anywhere in the world. This creates a solid foundation for reliable change detection.

How do you maintain calibration, consistency, and signal quality?

We treat calibration as a continuous discipline. It starts before launch. Each imager is fully characterized using traceable standards to establish a precise baseline. In orbit, calibration is ongoing. We continuously monitor performance, track drift, and cross-calibrate against trusted reference missions to maintain consistency across satellites and over time. The objective is not just accuracy, but stability and consistency over time.

Signal quality is also critical. We use a pushbroom architecture with time-delayed integration to maintain strong signal-to-noise. This allows us to detect subtle changes on the ground. Thermal behavior is another key factor. Small temperature variations can impact measurements, so we model, monitor, and correct for them continuously. If something does not meet standard, we fix it before launch. Long-term measurement integrity depends on getting it right upfront. All of this comes together in the ground segment, where data is aligned, corrected, and harmonized into a single, consistent data set. Calibration is not a step in the process. It is the process.

What can users expect as you move toward commercial operations?

They will see a shift from imagery to infrastructure. We are building a persistent, reliable data layer that allows corporations and governments to monitor assets, manage risk, and predict outcomes. Users will have access to daily global land coverage across visible, near-infrared, shortwave infrared, and thermal bands, acquired at a consistent local time with stable viewing geometry and precise geolocation. The data is delivered AI-ready. Calibration, atmospheric correction, cloud masking, and quality controls are applied so users can work directly with consistent measurements. Continuous coverage means there is no tasking over land required. Users are not requesting one-off images but accessing a persistent stream of measurement. This matters because instability downstream is costly. If data shifts after release, models break and trust erodes. We treat data release as a quality milestone. When it is delivered, it is stable and ready for use. Over time, the value compounds. The data set becomes a consistent record of change that supports monitoring, prediction, and decision-making.

What trends are you seeing in spectroscopy from space?

The biggest shift is from science to operations. Spectral data is no longer just for analysis. It is becoming an operational input for decision-making at scale. What changes is what matters. It is no longer about maximum spectral resolution alone. It is about calibration, consistency, and comparability across time, sensors, and conditions. Without that, the data cannot be trusted.

There is also a convergence between sensing and analytics. The expectation is AI-ready data that integrates directly into workflows and models. What is emerging are true measurement systems designed to support continuous monitoring, change detection, and prediction. That is where the market is going.

References

1. EarthDaily, Changed. Delivered Daily at Scale. EarthDaily.com. Available at: https://earthdaily.com/ (accessed 2026-04-03).
2. Wetzel, W. EarthDaily Releases First Processed Imagery from EDC-01. Spectroscopy. Available at: https://www.spectroscopyonline.com/view/earthdaily-releases-first-processed-imagery-from-edc-01 (accessed 2026-04-03).