How to Choose the Right Remote Sensing Service Provider
Selecting a remote sensing service provider in a market that is increasingly competitive is a decision that holds many technical and financial implications. The optimal choice is rarely going to be the provider with the largest satellite constellation or the most aggressive marketing, but rather the one whose technical capabilities and service model align most with the specific project at hand.
The range of raw data from public and commercial assets has altered the challenge from data acquisition to application-specific processing and analysis. This article attempts to present a framework for evaluating potential remote sensing partners so we can move beyond surface-level offerings.
The three pillars of remote sensing
Before engaging with any provider, an internal assessment is absolutely going to be key here. You should first determine if your need is data-centric or analysis-centric. A data-centric requirement focuses on the procurement of raw or pre-processed imagery with certain spatial, spectral and/or temporal resolutions. An analysis-centric need, though, requires a partner to develop and apply algorithms to extract specific information. The latter helps answer complex questions and generate actionable intelligence.
Next up is defining the project’s scalability and scope. Is this going to be a one-time analysis of a small study area, or a continental-scale, multi-year monitoring program requiring petabytes of data? The provider’s infrastructure must support your needs.
For ongoing programs, partnering with firms that offer custom processing pipelines that scale, such as DigitalSense, can be important when handling dynamic data streams. In the end, these partners can keep up with evolving analytical requirements.
Third, honestly evaluate your in-house expertise and whether you have a team of remote sensing scientists and GIS analysts ready to ingest and exploit Level-1 or Level-2 data. Or, do you require a full-service partnership that delivers end-product analytics and reports? This answer dictates the type of provider you need.
The full spectrum of service providers
Providers often specialize in distinct segments of the value chain – this isn’t a bad thing. Understanding these business models and the previous criteria is going to be key in shortlisting candidates.
At one end of the spectrum are the data and platform giants. Maxar Intelligence, for example, operates a very sophisticated constellation of high-resolution satellites, including its upcoming WorldView Legion. This service really does excel at providing premium, analysis-ready imagery and foundational data layers. This is for clients with pre-existing internal analytical capabilities, who simply need the data for themselves to analyze. Their value proposition is difficult to match in terms of providing access to high-quality and frequently refreshed raw data.
In the middle are the comprehensive software and platform providers like Esri. Through its ArcGIS ecosystem, Esri does a good job of empowering organizations to build their own geospatial capabilities. Their strength is not simply in data acquisition like Maxar Intelligence, but in providing an integrated suite of tools for clients to perform their own complex spatial analysis and data management. This is the choice for organizations investing in building and retaining in-house analytical supremacy.
Finally, specialized service firms like DigitalSense focus more on creating bespoke solutions for one-of-a-kind challenges. Rather than offering a platform for general use, they help develop custom processing pipelines and algorithms tailored to specific sensor data. So, this is particularly useful for firms with less experience in remote sensing. They can help with automated quality assurance or 3D reconstruction from non-standard imagery, for example.
Technical considerations for evaluation
A provider’s technical proficiency needs to be scrutinized. Sensor and data type expertise is, of course, non-negotiable. If your project involves SAR backscatter analysis for example, a provider specializing only in optical multispectral imagery is likely a poor fit. Probe their understanding of the specific sensor physics and collection geometries that will be relevant to your application. Their ability to source the right data is as important as their ability to process it.
Next, evaluate their processing capabilities. The quality of the final analysis is dependent on the rigor of their upstream processing chain. Ideally, you will ask for specifics on their approach to atmospheric correction and sensor-to-sensor normalization. A provider who can detail their expertise in building robust Level-1 satellite image processing pipelines is one who demonstrates a commitment to analytical integrity.
Finally, consider both the integration and delivery. How will you receive the output? Modern workflows demand more than a simple FTP drop. The availability of cloud-native delivery via APIs or cloud-optimized geoTIFFs means that there’s choice, and therefore, fit.
Making an informed decision
Choosing a remote sensing provider shouldn’t be rushed. A high-level assessment of project needs is needed, such as data versus analysis, scale, and the level of in-house skill. Then, it’s about assessing the provider’s experience within your project. While some companies provide the data, you need to fully accept who performs what analysis, and how it will be delivered.