The satellite industry is on the cusp of a revolution, with thousands of spacecraft operating simultaneously in low Earth orbit (LEO). This era of scale presents both opportunities and challenges, particularly when it comes to telemetry data. The sheer volume and complexity of telemetry data are pushing the industry towards what can be called the 'cardinality wall'.
Telemetry, once a manageable task, has become a distributed systems problem. Modern satellite buses expose tens of thousands of telemetry signals, streaming at sub-second intervals. This explosion of data creates high cardinality, where traditional ground system databases begin to break down. The challenge is no longer just collecting telemetry; it's preserving enough context and fidelity to operate autonomous, software-defined fleets at scale.
Many telemetry pipelines were designed for smaller missions and are struggling to support constellations generating millions of distinct telemetry streams in real-time. As fleets grow, these systems are failing in ways engineers didn't anticipate, pushing the industry towards the cardinality wall. The issue is not just about scaling; it's about preserving context and fidelity to ensure operational visibility, anomaly response, and mission resilience.
The problem compounds when operators try to retain telemetry long-term. Satellite programs store data for years or decades, requiring systems to support both real-time ingestion and large-scale historical analysis. This is a challenge few general-purpose databases handle well simultaneously. As cardinality and retention pressures compound, systems begin to fail, and operators are forced to simplify the data just to keep things running.
Loft Orbital, a company operating microsatellites in LEO, provides a real-world example of this challenge. As its platform scaled, it needed to handle millions of telemetry measurements per day, with ingestion rates reaching 10 million measurements every 10 minutes. Earlier approaches built on relational databases struggled to keep up with the volume and structure of the data, limiting visibility into system performance. By moving to a time series-oriented architecture, Loft was able to ingest high-frequency telemetry, maintain full context across missions, and deliver faster access to both real-time and historical data.
While removing tags, downsampling signals, or shortening retention windows may be a short-term fix, there's a dangerous tradeoff: loss of context. Context is crucial for engineers to correlate events across subsystems and for machine learning systems to predict component failures. Stripping away context makes anomaly detection significantly harder and can render the signals those models rely on disappear.
Breaking through the cardinality wall requires a shift in approach. Teams need to recognize when the current architecture has reached its limits and identify where cardinality is already impacting operations. This could manifest in delayed anomaly detection, slow replay of historical events, or gaps in data during peak ingest periods. Focusing on these pressure points provides a clearer path forward than attempting a full system overhaul.
Many teams are starting to decouple parts of the telemetry pipeline rather than replacing everything at once. Separating high-throughput ingestion from analytical workloads can stabilize real-time monitoring while longer-term changes are planned. It's also important to revisit strategies that depend on throwing away data to stay operational, as downsampling or stripping metadata may reduce load but also introduces blind spots that tend to surface later during anomaly investigations or design reviews.
Finally, telemetry systems need to behave more like distributed infrastructure than centralized databases. Data arrives out of order, in bursts, and from multiple locations. Systems that assume clean, sequential ingestion will continue to struggle in this reality. While there may not be a single fix for the cardinality problem, the teams that make progress are the ones that stop patching around legacy limitations and start isolating and redesigning the parts of their architecture that are already under strain.
In conclusion, the cardinality wall is a significant challenge for the satellite industry. However, by recognizing the issue, focusing on pressure points, and adopting a more distributed approach to telemetry systems, the industry can break through this bottleneck and ensure the resilience and operational visibility of future LEO missions.
