The spatial AI revolution: Entering the age of intelligence

The intriguing paradox about the information age is that it relies on semiconductor chips, which are fundamentally made from sand (silicon dioxide) — the most tangible and seemingly infinite resource on Earth. Yet, in 2023, the global digital storage capacity reached 110 zettabytes (110 followed by 21 zeros), which is a staggering figure; in fact, it is 15,000 times more than the number of grains of sand on Earth and it’s doubling every three years. The information age is suffering from excess information. Data is consuming the universe.

The velocity and quantity of information are overloading the ability to process it. This causes data-driven decision-making systems to fail. The limiting factor is human cognitive capacity to select, prepare and process the data, plus the ability to analyze it for meaningful insights. It is reminiscent of the early days of the Corona satellites of the TALENT KEYHOLE (KH) mission series that began in the 1950s during the height of the Cold War.

Understanding activities behind the Iron Curtain was critical for national security. The KH
satellites were expensive to launch and had short life spans. They used rolls of wet film dropped from space and captured by specialized aircraft with hooks to catch the canisters in mid-air. The low-resolution images (3 m to 5 m per pixel) were processed manually in darkrooms. Teams of 100 specialists, using razor knives and scotch tape, meticulously pieced together image strips into massive mosaics spanning several square meters. Working around the clock, assembling the full image would take up to five days, with subsequent analysis requiring another week. In total, from catching the film canister to delivering a final intelligence report, it took 17 days — a testament to imagery intelligence in the industrial era, characterized by massive operations demanding significant time and manpower, but it was too expensive and unsustainable.

“We live in a world where there is more and more
information, and less and less meaning.”

— Jean Baudrillard,

“Simulacra and Simulation,” 1994

In 1976, the technological landscape shifted dramatically with the launch of the KH-11 satellite, which could transmit 15 cm resolution images digitally to ground stations and was capable of distinguishing objects as small as a dinner plate. The satellite dramatically compressed intelligence-gathering timelines. Processing and analysis time decreased from 17 days to mere hours. The first digital image was shown to President Carter. That first image is believed to be of ongoing tensions in the Middle East, but it symbolized more than the triumph of technology; it represented a fundamental shift marking the end of the industrial era and ushered in the information age.

Advancements in imagery were paralleled by developments in mapping, driven by the need for accurate spatial referencing. Various technologies throughout the 1970s offered partial solutions, but a solution did not happen until 1981 when Esri introduced Arc/INFO, a breakthrough geographic information systems (GIS) software that could operate on minicomputers instead of huge mainframes. That formed the basis of modern spatial analysis and visualization technologies; coming together with digital imagery is what allowed the information age to overtake the industrial era.

In 2025, a similar technological transformation currently is underway. As the amount of information overwhelms existing systems, artificial intelligence (AI) is emerging as the solution. The information age is transforming into the intelligence age, where big processing meets big data. Advanced algorithms, machine learning and large language models (LLM) can swiftly and efficiently handle vast amounts of information. So, with data being the new oil, AI is the refinery.

The Esri Federal GIS Conference in February could have been promoted as the “Dawn of GeoAI” conference. The term Geo AI is a subset of Spatial AI, and it is in its infancy. Esri is incorporating AI into many of its applications. Companies at the expo were teasing Spatial AI solutions in their products and services.

What is Spatial AI?

When the transformative power of AI is combined with spatial information systems, magic happens. Value is created that did not exist before.

Spatial intelligence is the ability to think, visualize and understand in three dimensions. It is one of the primary types of intelligence. Currently, Spatial AI is capable of interacting with analysts using natural language to build models and perform tasks. Similar to so much else happening with AI, its capabilities are increasing rapidly.

With iterative learning, the AI repeats a task millions of times on various training data to perfect its abilities, running through different scenarios multiple times with different datasets while completing multiple tasks. The AI quickly learns and can eventually surpass humans. This makes AI a super tool.

Combine that capability with AI’s ability to access and infer an entire compendium of knowledge on a subject. The AI is able to ingest text, images, audio and video in minutes, and then reason and understand them all within the context of the parameters provided. Through its own AI agents, it will automatically run functions to garner insights, and then communicate those results through data visualizations, text, audio and natural speech. Spatial AI is an evolved form of AI able to understand data in the context of space and time within the body of knowledge it can access. It will monitor everything in real time to identify anomalies and hidden patterns and provide deep insights. It doesn’t just solve the information overload dilemma for data-driven decision-making, but it enhances it far beyond expectations.

The Coming World of AI Assistants

The future is already here. Reality is approaching science fiction at warp speed. A person living 100 years ago would only be able to understand the world of today as magic; and likewise, the world 20 years from now will appear magic to us.

Interfacing with a Spatial AI system is similar to the multi-dimensional world we already exist within. Flat screens, keyboard and mouse will be secondary tools behind natural language and natural gestures and immersive experiential environments. The Spatial AI- enabled world will blur the lines between what is virtual and what is real. Jobs, businesses and the economy already are transitioning. The most well capitalized businesses are investing in this new technology.

One of the industries at the forefront is healthcare. Imagine you are a neurosurgeon. Your patient has a glioblastoma identified by the MRI/CT scans uploaded into the Spatial AI Medical Assistant called SAIMA (pronounced Sāmă; when speaking with the system, you call it “Sammi”). The MRI/CT scans show a 3D model of the patient’s brain, highlighting the glioblastoma in red. Placing the integrated augmented reality (AR) glasses on, you can zoom in on the glioblastoma to see the extent of the growth and view it from any angle. This helps formulate a surgery plan. The patient’s medical records are in SAIMA along with the corpus of knowledge about glioblastomas. SAIMA is regularly updated with the latest algorithms and models. After reviewing the preliminary data, you have SAIMA run the spatial analytics and all the applied functions on the data. It takes approximately 35 minutes to complete. During that time, you review the SAIMA updates and go to lunch. You receive a text message from SAIMA after it completes its processing, letting you know it is finished without encountering any issues. SAIMA works with a system called VisAR, which is a precision surgical navigation system. After returning to your office, you put on the VisAR glasses to begin the review. Sammi begins by showing you the glioblastoma and pointing out it is a large, heterogeneous mass located in the frontal lobe and appears to be 4 cm to 5 cm in diameter, in an irregular shape with nodular and cystic components. As it goes through the review, it zooms in and rotates the 3D image, highlighting the exact area being talked about. You interrupt Sammi during this review and ask if the patient has been experiencing motor function issues since the tumor is in the frontal lobe, and you continue to probe further in a natural conversational tone as you delve deeper into the analysis. The conversation between you and Sammi is recorded and added to the file.

The review with Sammi takes several hours, during which a high-confidence surgery plan is developed that you will present to the multidisciplinary tumor board, who will further query SAIMA. This thorough process ensures the best results and further trains SAIMA about glioblastomas, which will be used for a post-surgery debrief and for insurance purposes. Following a successful board meeting, SAIMA proceeds to reserve the operating room, schedule the patient, and create a detailed surgery plan with specific duties and exact times for each member of the surgical team. This plan is then disseminated to all members of the surgical team and preoperative staff. A detailed surgical procedure file is generated, which serves as a navigation file, similar to Waze or Google Maps, providing step-by-step instructions to guide the surgery. This file will be loaded into ROSA (Robotized Surgical Assistant), a high-precision robotic surgeon.

On the day of the surgery, you wear special Bluetooth gloves that are synced with the SAIMA/VisAR glasses and ROSA. In real-time, magnified between 15x and 40x, you observe ROSA surgically removing the cancerous tissue. Overseeing the process, you see a tumor that has spread beyond the original CT/MRI scan and zoom-in on the tumor, and you take control of ROSA to manually remove the tissue. The surgical system uses a “differential engine” concept to scale down the surgeon’s movements to match the magnification level of the procedure, allowing for precise and delicate tissue removal. This means that the surgeon’s natural movements are reduced to a smaller, more precise scale, enabling accurate and intricate procedures. For example, a 1 cm movement by the surgeon might be translated into a 0.1 mm movement of the robotic arm, allowing for high-precision work. The system is dependent upon a high-level of spatial intelligence to make those calculations in real-time.

Afterward, you return the surgery back to the automated control of ROSA to follow the surgical procedure file plan. Throughout the fully immersive procedure, you speak with Sammi in a calm, natural language and responsive manner.

The patient, a married middle-aged father of two, not only survives but thrives because of the accurate analysis of SAIMA and the precision of ROSA, with you overseeing the entire process. The Spatial AI-based surgical system allows you to do what you wanted to do as a neurosurgeon and save people’s lives.

Nothing is Permanent Except Change

Breakthrough innovations, such as the internet, have changed the world. Spatial AI is going to do the same. These technologically driven schisms are huge opportunities. One can only speculate how it will alter the future. Once a technology takes hold, and it becomes obvious there is no going back, its adoption will accelerate, and in those moments, careers make exponential leaps. Those in front of it will make substantial gains. Position yourself accordingly. Learn about Spatial AI and Geo AI. Carve out your own specialty, such as Spatial AI/AR (augmented reality), Spatial AI/VR (virtual reality), Spatial AI/XR (mixed reality), and Spatial AI/FMV (full motion video). The future is yours to imagine.

WILLIAM TEWELOW is a designated Geographic Information Systems Professional. He has a master’s degree in Organizational Leadership with a focus on Performance Management, a bachelor’s degree in Intelligence Studies focused on geospatial intelligence, and an undergraduate degree in Geographic Information Technologies. William retired from the Federal Aviation Administration in 2025 after 16 years in various roles supporting geospatial information for aviation operations in the national airspace. He is a graduate of the management fellowship Program for Emerging Leaders where he served on special assignment to the Department of Transportation, leading a national strategic geospatial initiative under the authority of the White House Open Data Partnership.