Urban Context Analysis AI Agent Tutorial

In this video, I'll walk you through how to use AI GIS agents to quickly generate detailed site reports and mapping for any location. In this video, I'll walk you through how to use an AI agent to analyze a site, from urban context and 15-minute city assessment to environmental risks and residential real estate insights. Let's start with urban context analysis. First, go to ino.world and log into your workspace. If you use the new map function, you can generate both 2D and 3D mapping with AI like I showed in my previous video. You can find the link in the description below, but in this tutorial, we'll be using the new agent function. Here, we can either choose a pre-built AI agent template or write our own prompt from scratch. I'll be honest, I'm a bit lazy, so I just use the preset agent for urban context analysis. It automatically shows the prompts used for the site study, and we can always edit them or add additional features we want the AI to analyze. Here, I'll use the default settings first. For the location, you can either upload a CSV file with geolocation data or simply select the site directly on the map. For the study area, I'm using a radius-based buffer where you can customize the distance. You can also choose a custom site boundary, either draw it manually or define it in the CSV. Once everything is set, just click Generate. Then the AI will outline the methodology it's using for the urban analysis, covering everything from buildings and open spaces to street networks, design guidelines, and flood zones. Once the agent is activated, we can select our study site directly on the map. The data is globally accessible, so you can analyze any location. Since I have a research project in Chicago, I'll use that as my example and select the area I want to study. It takes about 5 minutes to run the full checklist, which covers 26 different analysis factors. Once it's done, the AI generates a comprehensive site analysis report. The AI also gives the site an overall score, along with a detailed explanation of how that score is calculated. For example, in this case, the score is mainly influenced by the scale context, which carries about 40% of the weight. The area is generally mid-rise, around 7 floors on average, but there's also a very tall outlier nearby, reaching up to 83 floors, which can affect how well a new design fits into the streetscape. It also breaks down the site into opportunities, constraints or risks, and a recommended integration strategy for the design. It also summarizes the data into clear numerical metrics, which is extremely useful for presentations. For example, it provides the number of buildings mapped, the median building height in floors, and the median building footprint in square meters. These are all key metrics for site analysis and reporting. It also generates charts to visualize the data, such as building height distribution, primary façade materials, activity levels, access, and traffic conditions. All of the key factors analyzed by the AI are clearly visualized. Normally, even if I use AI to gather the data, I still need extra time to turn it into presentation-ready graphics. But this AI agent combines everything in one place, which is really efficient. And finally, we can export the full report as a PDF for sharing or presentation. Beyond exporting a PDF, we can also sync the project to iKnow for further editing, modify the map, switch to 3D or Style 2D, and then export it as a PNG or PDF. In the classic iKnow editor, there are many pre-made map templates you can use, which saves a lot of time compared to Photoshop. Just like in QGIS, you can filter and categorize different data layers. For example, I turn on the building type layer, which lets me see different functions directly in the 3D map. In the second part, I'll test the 15-minute city assessment using AI. I'm still using the built-in model, just click the agent to activate it. This time, I'll try a city outside the US, so I update the location in the prompt and keep everything else the same. Then I activate the agent, review the methodology it's using, and select the site on the map. A few minutes later, just like the previous study, the AI generates a score for the location. This score is based on multiple factors that influence the 15-minute city assessment, such as access to amenities, schools, and transportation. The AI maps and evaluates all of these layers to give a clear understanding of accessibility. The AI agent evaluates how well a location provides daily services and amenities within a comfortable walking or cycling distance, typically within 15 minutes. It focuses on availability, accessibility, and overall neighborhood convenience. Based on this, it generates a 15-minute city score. For example, a score between 80 and 100 indicates a highly walkable area, where most daily needs can be met locally. I tested a suburban area in London, and as expected, the score is lower due to reduced access to amenities and services within a 15-minute range. Next, I use the AI agent for environmental risk analysis. This time, I write my own prompt and list all the key factors I want the AI to evaluate, such as flood risk, sea level rise projections, and social vulnerability. I also define the buffer radius for the study area. Based on this, the AI organizes the workflow into 33 different analysis factors. Then, I activate the agent to run the full assessment. This time, I select a suburban site in Tennessee, U.S. as my study area. After a few minutes, the AI extracts the data into layered maps, similar to GIS. Each factor is organized into its own layer, and every shape is clickable, so I can explore detailed information directly. I can also turn layers on and off depending on what I want to study or present. Then, the AI generates a full evaluation, including scores based on all 33 criteria. It also highlights what's driving the environmental risk, from pluvial flood sensitivity to vulnerable populations. The data is summarized into clear numbers and charts, such as a vulnerability index from 0 to 1, extracted from sources like the U.S. Census, which makes the results reliable and easy to use for analysis. For flood data, it pulls directly from FEMA. And for climate stressors, like rainfall intensity, it uses sources such as NOAA Atlas 14. It's pulling from the right datasets automatically, which saves a huge amount of time on research. One feature I really like is that it clearly lists all the data sources, so I can always go back and verify the reliability if needed. And finally, we can export the full analysis as a PDF report. Lastly, I want to show how to use the AI agent for residential real estate analysis. It's extremely helpful for real estate development, especially for site study and location selection. iKnowWorld also provides a preset AI agent specifically for residential real estate analysis, which makes the workflow very efficient. You can simply select the agent, and it will automatically generate a structured methodology and set of evaluation criteria for the study. The process is similar to what we've done before. Define the study area using a buffer radius, and then activate the agent. In this example, I choose a site in Dallas. Within a few minutes, the AI generates a development potential score, along with a detailed breakdown of the factors influencing that score. These factors include demographic data from the census, traffic conditions, income levels, accessibility, and surrounding amenities, all of which are critical for evaluating residential development potential. What's really powerful is that the AI not only gives a score, but also clearly explains why the site performs the way it does, making the results much easier to interpret and present. I can then run the same analysis on another site and directly compare the results, helping me identify which location has stronger development potential. This kind of workflow is extremely useful for early stage site selection, and it works well for both client presentations and in-house feasibility studies, where clear data-backed comparisons are essential. You can try the AI agent yourself on ino.world. I've added the link in the description below. If you have any questions, feel free to drop them in the comments. Thanks for watching, and I'll see you in the next video.