EOS Data Analytics

What is EOS Data Analytics?

EOS Data Analytics, abbreviated as EOSDA, is a company specializing in satellite intelligence and geospatial analysis; it was founded in 2015 by Max Polyakov and is headquartered in California, United States. The company combines its own EOS SAT-1 data, data from partner satellites, remote sensing algorithms, and industry expertise to create decision-making tools for use in agriculture, forestry, environmental management, finance, insurance, and infrastructure development.

EOSDA is not a single AI application; rather, it is a set of web platforms, mobile reconnaissance tools, development interfaces, and customized analysis services. Users can obtain information on vegetation, soil moisture, weather conditions, land changes, and risk indicators from satellite images, and use this data together with ground-based surveys to make informed business decisions.

Core products and current status

Main functions of EOSDA Crop Monitoring

Plot establishment and satellite observations

Users can draw the boundaries of farmland on the map, or upload existing boundary files and add information regarding crops, sowing, harvesting, and past growing seasons. The platform then matches these data with new satellite images, allowing users to view the changes in the same plot over time.

• Manage multiple plots and seasons, and keep centralized records of planting, irrigation, cultivation, and harvesting.
• Crop growth can be monitored using available images such as Sentinel-2 and PlanetScope; the resolution and update frequency depend on the data products used.
• Invalid scenes are reduced through the filtering of clouds, thin clouds, and shadows, but persistent cloud cover can still result in observation gaps.
• Switch between the global view and plot analysis to prioritize the identification of plots where there has been a significant decline in the vegetation index recently.

Vegetation index, soil moisture, and trends

The platform calculates multi-spectral bands into indicators such as NDVI, NDRE, MSAVI, RECI, and NDMI, which are used to assess vegetation vigor, early growth, and moisture status. These indices reflect remote sensing signals; they do not enable the direct detection of specific pests, diseases, or nutritional issues.

• On the timeline, compare different dates and various indices to observe anomalies, recoveries, and seasonal changes.
• Identify recurring patterns by combining historical vegetation and weather records to assist with crop rotation and resource planning.
• View the soil moisture products for the surface layer and root zone; the actual spatial resolution and update frequency depend on the data layer.
• Extended indices or customized band combinations are used to address specific crops and research questions, with some of these capabilities requiring additional services.

Risk maps, weather, and growing periods

Risk maps take into account recent changes in indices, weather conditions, and the growth stage of crops, thereby indicating potential problems such as diseases, extreme temperatures, strong winds, rainfall, or abnormalities in vegetation. The 14-day weather forecast, historical weather data, and indicators of heat and cold stress can help determine the best times for spraying, plowing, and monitoring.

• View the progress of crop development based on the BBCH growth stage and sowing date.
• Receive alerts regarding abnormal index values and weather risks to narrow down the areas that require on-site inspection.
• Connecting to ground-based weather stations to improve the availability of local data is possible, depending on the account settings.
• Weekly reports are used to summarize changes in the index, as well as variations in precipitation and temperature.

Variable application maps and yield estimation

Variable application maps divide management areas based on vegetation, productivity, altitude, or agricultural machinery data, to serve as a guide for seeding, fertilization, irrigation, and pest control activities. Users can export these maps and hand them over to compatible devices or teams for use, but the amounts of pesticides to be applied still need to be determined by agronomists taking into account soil samples and local regulations.

Yield estimates provide short-term forecasts by taking into account factors such as crop variety, irrigation, and weather conditions, and they are useful for planning resources and logistics. The results of the model are influenced by the quality of the input data, regional calibration, extreme weather conditions, and actual field management practices; therefore, they cannot be considered a guarantee of actual yields.

Reconnaissance, operations, and team management

Once an abnormal area is identified, managers can create reconnaissance tasks on the map and assign personnel to carry them out. The mobile app allows users to navigate to the target location, take photos and make observations offline; once the network connection is restored, the reports can be synced to the team account.

• Keep logs of activities such as cultivation, irrigation, fertilization, spraying, sampling, and reconnaissance.
• Share plots, indices, tasks, and reports by role to reduce redundant communication.
• Create automated or custom reports to display the status of plots to management, customers, and partners.
• Synchronize boundary and some agricultural machinery operation data via the John Deere Operations Center.

Main functions of EOSDA LandViewer

LandViewer is intended for users who need to search for and process Earth observation data on their own. Users first define the area of interest and the date, then choose between optical, multispectral, radar, or commercial high-resolution data sources for previewing, processing, and downloading.

• Search for available scenes from Sentinel, Landsat, EOS SAT-1, and the directories of partner data providers.
• Use natural colors, pseudocolors, vegetation indices, and custom band combinations.
• Compare forests, farmlands, coasts, cities, and disaster areas using time series and change detection.
• Calculate regional statistics, partitions, and raster metrics, and export the results as JPEG, PNG, GeoTIFF, or vector formats according to the plan.
• Preview high-resolution commercial images and purchase them by area or by scene; the minimum order amount and licensing requirements are determined by the data provider.

Free or subscription-based access to low- and medium-resolution images does not mean that commercial images are also included in the package. It is necessary to verify the data provider’s licensing terms, the minimum area required, the rules regarding re-access, the level of cloud cover, and the permitted ways of redistributing the images before making a purchase.

EOS RayVision and custom analysis

EOS RayVision is designed for organizations that need to monitor large areas, as well as any changes or impacts in those areas. Customers define the areas of interest and the objects to be detected, while EOSDA provides structured data, analysis reports, and interactive maps. It can be used for inspecting infrastructure, assessing the impact of disasters, monitoring land use, forests, mining activities, maritime operations, financial risks, and ensuring security.

RayVision is not a generic generative AI that can produce results simply by entering a prompt; for this service it is necessary to specify the area, resolution, sensor type, frequency of revisits, verification method, delivery format, and deadline. The price is determined by the sales team based on these specifications.

API Connect feature

EOSDA API Connect offers REST interfaces that allow developers to integrate functions such as satellite search, image processing, statistics, rendering, and weather data into agricultural software, GIS systems, insurance models, or internal analysis pipelines. Once an account is activated, an API Key is provided, which must be included in the requests sent.

API integration steps

1. Create an API Connect account and request a trial activation or a paid plan from sales.
2. Retrieve the personal API Key from the user’s backend; do not include it in public code repositories or client packages.
3. According to the document, select the search, imaging, statistics, rendering, weather, or zoning interface.
4. Use the test area to verify the coordinate system, date, cloud cover, bands, indices, and output format.
5. Record the number of requests required for each business process, and implement queuing, caching, and retrying based on account limits.
6. Compare the remote sensing results with ground samples to confirm that the required accuracy has been achieved, and then expand the area.

The development documentation provides examples in Python, JavaScript, and Node.js, as well as Postman resources; however, no officially released SDK package was identified this time. Code examples are not equivalent to an SDK, nor do they indicate that the EOSDA core platform is open source.

Use workflows

1. First, determine the objects to be monitored, the area involved, the date range, the acceptable resolution, and the frequency of updates.
2. Agricultural users draw or upload plots in Crop Monitoring, while image analysis users define areas of interest in LandViewer.
3. Filter satellite scenes and check the shooting time, cloud cover, spatial resolution, and data licensing.
4. Choose indices that are suitable for the growth period or specific operational issues; do not rely solely on a single NDVI value.
5. Anomalies are identified by comparing historical time series, weather conditions, and adjacent plots, in order to rule out changes caused by cloud shadows, bare soil, and harvesting.
6. Convert suspected risks into reconnaissance tasks or additional imaging requests, which are then verified through on-site photos, sensors, or professional experts.
7. Generate reports, data layers, or API results, while retaining the scenario, algorithm, date, and decision-making process.

Suitable for users and scenarios

• Farms and agricultural groups: remotely monitor large areas of land, arrange inspections, and manage variable inputs.
• Agronomic consultants and cooperatives: managing plots belonging to multiple clients, generating reports, and providing white-label services.
• Banks and insurance companies: Use historical growth trends, weather conditions, and the impact of disasters to assist in due diligence and claim verification.
• Forestry and Compliance Team: Monitors logging, restoration, canopy health, fire damage, and land changes.
• Governments, research institutions, and non-profit organizations: carry out projects related to food security, the environment, disasters, and land use.
• Developers and geographic information teams: Create applications in the fields of agriculture, meteorology, and remote sensing using APIs.
• Energy, Mining and Infrastructure Agency: Uses RayVision to continuously monitor changes and risks surrounding its assets.

Prices and billing

EOSDA’s various products use different billing methods, and they cannot all be categorized under the label of \"free with premium features\". Crop Monitoring is billed based on area and the scope of the service; LandViewer is charged through subscriptions and purchases of images; the API is billed per request package, while RayVision and custom projects are priced according to quotes provided by sales.

The example prices listed for LandViewer may vary depending on the settlement period, region, taxes and fees, as well as the specific products; high-resolution images are also affected by quotes from third parties. Before making a purchase, it is necessary to refer to the settlement page in the account, the data order, and the sales contract.

Notes on renewal and refunds

• The monthly or annual subscription to LandViewer may be automatically renewed, and it should be canceled from the account before the next billing date.
• Subscription fees paid before the cancellation takes effect are generally not refunded; they can still be used until the end of the current period after cancellation.
• The rules regarding refunds and termination for Crop Monitoring, APIs, and enterprise projects may be specified separately in individual orders.
• Commercial imaging orders come with requirements regarding data providers, minimum area, and licensing conditions; it is necessary to check whether they can be canceled or modified before submitting them.

Platforms and integration

API, SDK, GitHub, and open-source status

• EOSDA API Connect provides public REST documentation and an API Key authentication process.
• To try it out, you need to contact sales for activation; the documentation states that each user on a trial account is entitled to 1,000 requests.
• The document examples illustrate request rate control; the actual restrictions are determined by the account plan and the interface responses.
• No officially released SDK packages or public GitHub repositories that represent the core product were identified this time.
• The terms state that the product may contain third-party open-source components used under their respective licenses; this does not mean that the EOSDA product itself is open source.
• The core platform, remote sensing algorithms, and AI models are proprietary services; it is not permitted to reverse-engineer them, collect data from them in bulk, or resell access rights to them.

Privacy and security

EOSDA collects registration and contact information, communication records, service usage data, search history, IP addresses, browser details, device information, location data, and Cookie data; it may use this data for purposes such as authentication, support, analysis, marketing, and security. Corporate locations, geographic coordinates, and business activities can also be considered commercially sensitive, so it is necessary to clarify the scope of data processing at the time of procurement.

• The security instructions specify AES-256 encryption for static data and TLS encryption for the transmission process.
• The platform employs role-based access control, and it provides information on security audits, vulnerability assessments, incident response, logging, and patching processes.
• Crop Monitoring and LandViewer have been listed as software that has passed the AWS infrastructure assessment.
• Privacy requests may relate to access, correction, deletion, and rights applicable in certain regions, but the retention period depends on business, contractual, and legal obligations.
• Disabling some cookies can affect functionality, and the service may also use third-party analytics tools.
• API Keys, precise farm boundaries, customer lists, and the locations of confidential assets should not be included in public code or screenshots.

Copyright, data licensing, and commercial boundaries

EOSDA permits use for personal, consumer, or internal business purposes within the terms set forth, but it prohibits impersonating its services, reselling access rights, or extracting data without authorization. The satellite images in LandViewer may belong to EOSDA, public satellite projects, or commercial data providers; the rights to copy, display, derive from, and redistribute such images vary depending on the context.

• Before downloading, check the signature of the specific image and data provider, as well as the conditions for commercial use and redistribution.
• Customer reports can be used for internal decision-making, but they do not constitute the original images or platform interfaces as data that can be sold as one’s own.
• When the API output is used in customer products, the caching, storage, display methods, number of users, and scope of sub-authorization must be specified in the contract.
• Third-party base maps and geocoding in the map may also be subject to the corresponding service terms.
• For regulatory, insurance, or judicial purposes, the original scenario, time, processing parameters, and verification records must be preserved.

Main advantages

• By combining its own agricultural satellites with various cooperative constellations, it is possible to select optical, radar, and commercial imagery depending on the task at hand.
• A relatively complete remote sensing workflow, ranging from image search and index calculation to field operations and report generation.
• Crop Monitoring is suitable for agricultural users who do not have a professional remote sensing team, while LandViewer offers more control over the analysis processes.
• The API covers search, imaging, statistics, rendering, and weather, and is suitable for integration into existing GIS and agricultural systems.
• RayVision can deliver complex projects as data layers, reports, and interactive maps.

Capacity limits and risks

• Satellites cannot penetrate all clouds, and optical images are affected by clouds, shadows, smoke, as well as the timing of re-visits.
• Spatial resolution determines the size of visible objects; a low resolution is not suitable for identifying individual plants or very small features.
• Abnormalities in vegetation indices can only indicate potential risks; they cannot confirm the presence of pests and diseases, nutrient deficiencies, or yield issues in the absence of field evidence.
• Weather, yield, classification, and change detection are all results of the model; regional migration can introduce errors.
• Different products, additional features, commercial images, and APIs are billed separately; the overall budget needs to be estimated based on the workflow.
• The independent forestry platform is no longer in use; current users should choose LandViewer or consult about customized solutions.
• Sensitive geographic location data, infrastructure information, and farm operation data require strict access controls and contractual oversight.

Purchasing and evaluation recommendations

1. Select a small number of representative plots or areas with ground-based records, and define measurable verification objectives.
2. Compare the impact of different satellites, resolutions, cloud cover, and update frequencies on actual decision-making.
3. Record the costs associated with video recordings, API requests, manual verification, and reporting required to complete a business transaction.
4. Check data permissions, retention periods, team privileges, API restrictions, as well as rules for renewal and cancellation.
5. Before expanding procurement, assess false positives, false negatives, and regional suitability through on-site investigations or authoritative data.

Frequently Asked Questions

Is EOS Data Analytics a generative AI tool?

It is not a general-purpose generative AI designed primarily for creating images or text. Instead, it uses AI, machine learning, and remote sensing algorithms to analyze real satellite and weather data, providing spatial decision-making information for agriculture, forestry, and various other industries.

Can EOSDA Crop Monitoring be used for free?

The product offers a trial version, but the current payment options are Essential, Professional, and Enterprise. The exact scope of the free version as well as the relevant pricing details can be found on the account pricing page or in the sales quotation.

What packages are available for LandViewer currently?

The available plans are listed as the free Starter plan, the Explorer plan at 19.99 dollars per month, and the Pro plan at 49.99 dollars per month. High-resolution commercial images are usually purchased separately based on area or scene; the final price is indicated on the settlement page.

Is there a free trial for the EOSDA API?

A trial version can be requested; the documentation indicates that the trial key allows for 1,000 requests and must be activated by the sales team. The paid version is priced based on the number of requests, and the exact price in dollars has not yet been determined.

Does EOSDA provide SDKs or open-source code?

The REST documentation, Python and Node.js examples, as well as Postman resources have been verified, but the official SDK package has not yet been confirmed. The inclusion of third-party open-source components in the product does not mean that the core platform, data, or algorithms are open source.

Is it possible to determine crop diseases just by looking at satellite images?

It’s not possible. Index and risk maps are useful for identifying abnormal areas, but the specific reasons for these abnormalities need to be verified by taking into account weather conditions, on-site inspections, as well as knowledge of soil and crops.

Is EOSDA Forest Monitoring still in operation?

The original standalone platform has been discontinued; the relevant forest monitoring capabilities are now available through LandViewer and customized forestry solutions. The old version should not be considered a separate product that can still be registered directly.

Does EOSDA have a mobile app?

It has been confirmed that the Android version of the Crop Monitoring app can carry out reconnaissance tasks, record information on-site, and generate synchronized reports. The iOS version has not been separately verified on the store page; it is necessary to check the developer’s name when downloading it.