The Grid Installation and Space Reuse / Ad Hoc Practice © Trieu Chien**

Behind layers of plaster, paint, and finishes lies an intricate network of pipes, electrical conduits, beams, and other structural elements that make a building function and stand, yet remain invisible to the everyday eye. Within these layers, traces of different periods accumulate: replaced systems, improvised adaptations, and technical solutions that once responded to specific contexts and urgencies. In adaptive reuse, the greatest challenge often begins before construction even starts, which is understanding what lies within when little or no reliable documentation exists. During a renovation, pleasant or unpleasant surprises are inevitable. The unexpected is part of the process, but it also represents cost, delay, and risk factors that often discourage investors and professionals from engaging in this type of project.

Powerhouse Arts / Herzog & de Meuron + PBDW Architects © Iwan Baan

It is precisely this “invisible” dimension that the latest research by global design and make software company Autodesk and design and engineering firm Arcadis seeks to address. By combining artificial intelligence, sensors, and predictive modeling, the teams are developing an experimental tool capable of “seeing through walls.” This system uses multiple sources of information, such as laser scans, sensors, old floor plans, and geospatial data, to create intelligent three-dimensional models of existing buildings. From these models, AI can infer and predict what is not directly visible, including the position of pipes, the condition of materials, or the structural health of certain components.

According to David Benjamin, Director of AEC Industry Futures at Autodesk, the goal is to offer architects and engineers “a kind of x-ray vision for buildings.” As he explains, “Laser scans are great because they can tell you about all of the visible aspects of an existing building. But they can’t tell you about the invisible aspects, including what is inside the walls and the condition of each material. If we can create a kind of x-ray vision for buildings, we may be able to unlock new possibilities for material reuse, circularity, and decarbonization.”

Rather than relying exclusively on detailed surveys, inspections, or invasive investigations during construction, the idea is to triangulate diverse sets of data to reveal hidden patterns and produce more accurate predictions about a building’s real condition.

This research involves applying AI to multiple data sets that are limited instead of one data set that is comprehensive. When we combine imperfect data from old floor plans, laser scans, sensors, and GIS information, we can derive better insights than if we had a larger and more pristine data set from a single category. This research explores how AI can leverage existing things. Instead of generating new designs, we are teaching AI to interpret and connect what already exists. — David Benjamin, Autodesk Research

Reuse as a Cultural and Environmental Strategy

For Mansoor Kazerouni, Global Director of Architecture and Urbanism at Arcadis, retrofit is more than a technical exercise; it is a cultural and ecological responsibility. “Reusing existing structures remains one of the most effective ways to reduce carbon emissions and preserve the character of our cities.” He explains that one of the main challenges in renovation projects is precisely “the unknown.” “Existing conditions aren’t always clearly documented. You don’t truly know what you’ll encounter until work begins.” Outdated drawings, undocumented alterations, and hidden systems often lead to unexpected discoveries that affect costs and schedules. Technologies that can “see through walls” could completely change this dynamic.

Screenshot of the Video

Seeing what’s behind the walls before work begins can transform on-site decision-making. It builds team confidence, reduces surprises, and allows repairs or upgrades to be planned with precision. — Mansoor Kazerouni, Arcadis

In practice, this means less waste, more accurate budgeting, and safer construction sites. Both experts highlight that the future of sustainable architecture depends directly on understanding what already exists. With early access to information about structures, systems, and materials, teams can plan interventions more precisely, avoiding surprises and rework. This predictability increases confidence and allows decisions to be guided by data rather than reactions to unforeseen conditions. Knowing the exact location and condition of systems facilitates the reuse of components, reduces waste, and optimizes resources. Beyond improving design and construction efficiency, this integrated vision could contribute to more realistic budgets, more controlled timelines, and safer workplaces.

Buildings account for 40% of global carbon emissions. If we can make it faster and easier to reuse existing buildings, we may be able to make a real impact despite the immense scale and the short timeframe involved in the carbon problem. — David Benjamin

Arcadis has significant experience in this field. The renovation of Castellana 66, in Madrid, is a prime example. The 1990 office building underwent a comprehensive sustainability plan that turned it into one of the most energy-efficient buildings in Europe. The intervention preserved the original structure, prioritizing embodied carbon retention and a complete upgrade of environmental performance. The project achieved a 55% reduction in annual energy use and prevented the emission of 10,800 tons of CO₂ into the atmosphere.

Screenshot of the Video

Toward Smarter and More Sensitive Renovations

More than precision, Autodesk’s research also explores uncertainty, an essential yet often invisible component of AI-driven decision-making. “Every time AI makes a prediction, it has a sense of how certain or uncertain it is, but this information is rarely visible to the user,” explains Benjamin. “We think it would be helpful for decision-making if a user could see both an AI prediction and a level of uncertainty at the same time.”

To reuse is also to reveal, and to remain open to surprises. Tools like these represent not only a technological leap but also an epistemological shift in how architects read, interpret, and intervene in the built environment. By transforming uncertainty into knowledge, they make the invisible visible and enable a more informed form of reuse that bridges research and practice, precision and sensitivity, innovation and preservation. If we take to heart the words of Carl Elefante, former president of the American Institute of Architects, who famously said that “The greenest building is the one that is already built,” the future of architecture may depend on how well we learn to see what is already in front of us.

Arcadis recently restored the Brooklyn Paramount Theater to its former glory. Image Courtesy of Arcadis

***This article is part of the ArchDaily Topics: Building Less: Rethink, Reuse, Renovate, Repurpose, proudly presented by Schindler Group.****Repurposing sits at the nexus of sustainability and innovation — two values central to the Schindler Group. By championing this topic, we aim to encourage dialogue around the benefits of reusing the existing. We believe that preserving existing structures is one of the many ingredients to a more sustainable city. This commitment aligns with our net zero by 2040 ambitions and our corporate purpose of enhancing quality of life in urban environments.*Every month we explore a topic in-depth through articles, interviews, news, and architecture projects. We invite you to learn more about our ArchDaily Topics. And, as always, at ArchDaily we welcome the contributions of our readers; if you want to submit an article or project, contact us.

**Cite: **Eduardo Souza. “See Through Walls: Adaptive Reuse Through Data, AI, and Circular Design” 11 Nov 2025. ArchDaily. Accessed . <https://www.archdaily.com/1035859/see-through-walls-adaptive-reuse-through-data-ai-and-circular-design> ISSN 0719-8884