AI three ways

It should not come as any particular surprise that Esri has integrated artificial intelligence (AI) capabilities into its leading geographic information system (GIS) technology, ArcGIS. Ever since the company was founded in 1969, it has been constantly innovating. Esri pioneered the development of GIS technology, and its continuing commitment to research and development keeps it at the forefront of the industry. But what are the AI capabilities in ArcGIS and what benefits do they offer?

Esri has enriched its ArcGIS platform with a suite of AI tools, models, workflows and agents, which, together, are commonly referred to as Geospatial AI or simply GeoAI. The approach centres on delivering three complementary AI capabilities.

AI-Powered Analysis 

The first is the analytical layer of tools and models. More than one hundred ready-trained models are available through the ArcGIS Living Atlas for tasks such as object detection, feature extraction and change analysis. These models allow organisations to classify land cover, identify assets from imagery or monitor environmental change without first assembling a specialist data science team.

Alongside these capabilities sits a newer generation of technology. Foundation models are trained on large and diverse datasets and can support multiple tasks rather than a single purpose. Esri is bringing both open-source and proprietary foundation models into ArcGIS together with location embeddings, creating new ways to reason about spatial information.        

AI-Assisted Workflows

The second capability is the growing set of assistants embedded directly within ArcGIS applications. The Arcade Assistant can generate and troubleshoot expressions across Map Viewer, Web Editor, Field Maps Designer and Experience Builder. The Notebooks Assistant helps users write and debug Python code, while the ArcGIS Pro Assistant, currently in beta, supports documentation queries, workflow guidance and code generation.

One capability I would particularly like to highlight is the Item Details Assistant. It can automatically suggest titles, summaries, descriptions, tags and field properties as content is published. This may sound like routine housekeeping, but metadata quality increasingly

determines whether information can be discovered, understood and trusted, both by people and by AI systems. Organisations preparing for AI should view metadata as a strategic asset rather than an administrative task.

Agentic AI

Beyond analytical models and workflow assistants sits a third development, agentic AI. Within ArcGIS, Esri is introducing agents capable of carrying out routine tasks and supporting more complex workflows. At the same time, Esri is embracing interoperability through the Model Context Protocol (MCP), allowing external agents built on other AI platforms to access ArcGIS tools and authoritative spatial data through standard interfaces.

In my opinion, this second direction is especially significant. It means an enterprise AI agent, regardless of where it originates, can access authoritative geographic information when asked to assess risk, evaluate service impacts or analyse movement patterns. It also reinforces the importance of metadata and data preparation. An AI agent can only discover, understand and use information that has been properly described and made accessible.

While the benefits of using GeoAI will inevitably vary team-by-team and company-by-company, depending on how it is used, there are three stand-out benefits that most organisations can aspire to gain.

Time savings

GIS professionals can use GeoAI to simplify routine analysis tasks and make helpful shortcuts. For example, they can use the new AI assistant in ArcGIS Pro create visualisations of complex data, using native language instructions like, ‘show me in a bar chart’. They can automate complex data analyses, evaluate multiple scenarios and find optimal solutions more quickly, particularly when working at scale.

Take the energy sector for instance. When undertaking complex site selection processes for new renewable energy schemes, project managers can use GeoAI to automate the analysis of solar radiation projections, wind-speed, subsurface conditions, grid capacity and other factors to identify potential risks and compliance considerations more quickly.

Improved foresight

The use of GeoAI can help organisations to detect anomalies amongst ‘big data’ and identify patterns that would have been almost impossible to spot manually. Taking this one step further, organisations can derive greater insight from their data to predict trends and model future scenarios.

In the infrastructure sector, for example, organisations can use GeoAI to analyse road networks, drone imagery and real-time traffic data to better understand the condition of road surfaces. They can then predict maintenance needs without expensive, time-consuming ground surveys and plan works more proactively to reduce disruption for road users

Optimal performance

AI capabilities can be built into organisational workflows to help improve staff productivity and business performance. Field workers, for instance, can use AI-enabled capabilities in ArcGIS to recognise features in photographs and automatically populate forms, enabling field workers to complete more surveys in a day.

To provide a different example, teams working in disaster response can analyse real-time drone and satellite imagery with GeoAI to get a faster understanding of an unexpected scenario, such as an earthquake. They can then communicates the extent of the crisis with real-time smart maps and plan the most appropriate emergency response for the areas worst affected.

Esri has been working to bring the fields of GIS and AI together ever since 2018. Even before that, as far back as 2009, it was introducing machine learning capabilities into ArcGIS. As this focus on innovation continues, more new AI functionality will be added to ArcGIS in the future. In the meantime, three categories of cutting-edge GeoAI capabilities are already available to you, along with three potential benefits that are too important to ignore.