The intersection of GIS and Machine Learning is evolving and bringing new use cases and applications of ML to the fore. These applications, spanning both the private and public sectors, are powered by large volumes of data captured by satellites, drones, cameras, LIDAR sensors, and more, all of which come together to provide a comprehensive view of the world. The sheer volume and variety of data create complexity in its management and usage.
As the applications of ML for GIS grow in complexity, generating high-quality ground truth or training data for these novel applications can become difficult. Although it is widely known that training datasets often need to be quite large, it is less known that these datasets increasingly need to be labeled by either subject matter experts or trained personnel in a variety of different fields.
Feature extraction or the process of extracting similar spectral, spatial, and texture attributes from geospatial imagery to power different use cases is foundational in dataset creation.
There is a unique opportunity today to create GIS training data utilizing data labelers and/or GIS Technicians that are trained and managed closely. With the range of applications within the field, a deep understanding of the type of the data required for each use case as well as subject matter expertise or a hybrid of the two, is required.
Geospatial intelligence for private and public sector
Geospatial intelligence provides geographical information and distribution of elements in a geographic space and is now an essential tool for everything, from national security to land use and planning to agriculture and a host of commercial and government functions.
The use cases of geospatial applications run a broad gamut of public and private sector activities, including land use planning, commercial and residential insurance, agriculture, national security, oil and gas exploration, and retail.
Let’s take the case of insurance companies. GIS can provide them precise location-based insights, which they need and can be used for risk management. Location-related information like where assets are situated, their closeness from hazards such as industrial areas, natural elements ,is important for insurance firms to develop risk profiles. Access to such information could be valuable for insurers to make informed decisions.
Similarly, in agriculture, geospatial intelligence can complement farmers’ efforts by providing them a bird’s eye view of fields and crops. This data is useful in understanding the spread of the yield, crop health, threats or availability of natural resources such as water bodies. All this can enable farmers or relevant businesses to take the right decisions for improved yields, and reduced time and effort.
In the public sector, there is tremendous innovation in geospatial intelligence. Often, defense departments of countries like the US use geographical data to assess security measures and deploy intelligent military operations. Companies like Maxar Technologies are providing ground imagery of the Russia-Ukraine conflict, in real-time, to share information about the Russian advance, to the world.
Likewise, defense departments across the globe use geospatial data and remote sensing to monitor enemy movement on ground or in the air – detection of unidentified aircraft, spy drones, fighter jets, etc.
Beyond this, geospatial is also being used by governments to keep an eye on possible natural disasters such as floods and earthquakes. This is important for planning and strategizing rescue missions swiftly to reduce loss of lives and properties.
When faced with this array of applications, workflows that help with scaling skilled teams, providing onboarding training, project management, and quality assurance throughout the project, must be built keeping in mind the final deployment of the data and the ML model.
Tools and techniques for GIS
For geospatial intelligence, data is gathered through satellites, drones and other aerial sources that capture everything within a specific geographic area, and the data annotation required varies depending on the final use case. At iMerit, workflows and project design are tailored to adapt to the vast range of geospatial applications.
For example, in the case of insurance companies, iMerit uses image classification and 2D polygon annotation to capture features of a building such as windows, doors, garages, swimming pools to assess the insurance premium rate.
In the case of the military, the data capturing is sophisticated and involves multiple methods and technological support. Technologies like SONAR are also used to gather data. The LIDAR method of data creates top-of-the-house images that can be used to create topographic maps, showing different elevations and spraying elevation models for different use cases.
In case of infrastructure inspection, drones are sent out with RGB video as well as LIDAR sensors to get very high-fidelity scans or models of infrastructure using which you can assess different areas of weakness or those that need to be replaced.
For example, California is known to have an aging energy infrastructure for power lines that have been updated with the help of geospatial intelligence. Texas is an example where snowstorm-affected infrastructure is being identified using drones and other aerial means and are being restored.
Future of data solutions for geospatial intelligence
The amount of data available and being collected has increased drastically. So has the demand for higher resolution and better-quality data. There are different types of data that are emerging, such as synthetic aperture data (SAR), which is relatively new. We are starting to see different ways of collecting data. The one that is really transformative is LIDAR, which is helping to create highly accurate 3D representations or models of different aspects of the world.
There is another interesting perspective which is increasingly being used, applying machine learning and computer vision to these data types to automate some of this analysis, which was typically done manually by GIS analysts.
GIS and machine learning will continue to evolve, bringing unique use cases and subsequent complexities in data usage to the forefront. This dynamism requires training and developing skills of data annotators to seamlessly provide high-quality data to feed artificial intelligence- and machine learning-led systems. It is thus vital for data annotation providers to bring together technology, talent, and technique to deliver, simplified, high-quality data for businesses and society to make informed decisions.
