Monitoring construction activity is crucial for businesses involved in construction because such activities can be time consuming and costly for companies and public institutions.
A recent platform, ConsTRACK, enables Earth Observation (EO) satellite data to be assessed in a user-friendly format that also shows updates and changes to sites during the progress of given projects.
The tool could be used to save money in remote monitoring of construction while it also shows applicability in other areas.
What is ConsTRACK?
The ConsTRACK tool has been created to help businesses and construction to be more efficient and also better utilize devoted resources. Construction can be costly for businesses as well as governments and taxpayers that commission major efforts.
Monitoring activities can be time consuming and also costly if major projects are done in remote or less accessible regions. Progress, at a day-to-day level, may also be difficult to observe, particularly if construction sites are large or spread over a large area.
Furthermore, with the use of contractors and multiple companies for any given project, the task becomes even more complex and relevant to monitor.
The tool applies an algorithm whereby new data, as they are acquired, are compared to previous imagery, allowing changes from the previous image to automatically be shown and provided for users. Recently, the tool has been seen as providing benefit for other areas outside of construction, including for monitoring related to waste water management, oil and gas, insurance, and water.
Using Satellite Data to Monitor Construction Activity
The types of satellite data that have been used with the tool not only include normal optical data but also thermal, multi-spectral, and air quality data. The types of satellites used are Sentinel-1, Sentinel-2, Sentinel-3, and Sentinel-5P.
Analyses could include coherence analysis from SAR data, deriving permanent scatters in landscapes and classifying features, including any change. Results can be fine tuned to detect specific changes such as construction change or air quality change in analyzing relevant imagery.
Rapid Acquisition Data to Monitor Development
While the ConsTRACK tool is a relatively recent development, this type of tool does highlight the increasing trend in using more rapid data acquisition to provide automated information to users that not only provides the data but also the analytical result of that assessed data so that it can be used appropriately as soon as information is obtained.
What is different in ConsTRACK is it deploys classification and assesses changes to classes and features as multiple data from different periods are assessed.
Tools to Map and Classify Features
While we are now seeing more recent applications and tools to help identify and classify features, including in construction and land use management, to expand capabilities, data repositories of relevant imagery and classified data sets used for training identification models will need to be created at larger scales to enable even more fields to benefit from such automated tools for analysis and classification. This includes tools that identify change between imagery such as ConsTRACK.
The methods are now well established among practitioners and industry, and in coming years the data will likely catch-up.
Tools such as ConsTRACK are promising in that not only do they do the automated assessment of landscape features but they can detect change and provide information of relevant change as soon as data are available to relevant users.
The tool also shows the next step in moving from simply classifying to also showing how features changes over time, including not only in optical form but also less visible forms such as air quality. In coming years, more similar capabilities in other areas in land use and landscape monitoring are likely.
 For more on analytical capabilities from a recent article, see: https://eo4society.esa.int/2021/09/29/a-tool-for-artificial-surfaces-change-tracking-with-eo-data-constrack/.
 For more on a tool to do automated analysis and identification of smoke, see: Larsen, A.; Hanigan, I.; Reich, B.J.; Qin, Y.; Cope, M.; Morgan, G.; Rappold, A.G. A Deep Learning Approach to Identify Smoke Plumes in Satellite Imagery in Near-Real Time for Health Risk Communication. J Expo Sci Environ Epidemiol 2021, 31, 170–176, doi:10.1038/s41370-020-0246-y.
 For more on irrigation works and identification, see: Saraiva, M.; Protas, É.; Salgado, M.; Souza, C. Automatic Mapping of Center Pivot Irrigation Systems from Satellite Images Using Deep Learning. Remote Sensing 2020, 12, 558, doi:10.3390/rs12030558.