International architecture journal awards Tec professor for best paper

Tec Guadalajara professor Luis Palomares has been recognized for publishing the best paper of 2025 in the journal Architectural Intelligence. This paper discusses his research on walkability and analyzes how urban space is experienced in cities. 

The publication came about after he was permitted to give a virtual presentation at the International Conference on Computational Design and Robotic Fabrication, held in Shanghai, China, on the topic of Trans-Individual Intelligence (intelligence beyond the individual). 

Palomares’s presentation: Enhancing graph machine learning work flows with visual programming for urban design, was based on his thesis about the use of artificial intelligence (AI) for urban issues and architecture. 

This event was organized by Tongji University and Digital Futures, an organization focused on technology-based architecture and urbanism, with the journal Architectural Intelligence forming part of the conference.  

According to the international publication, the paper received the award as a result of its: 

Following the presentation, the organizers invited the academic to expand his research on walkability in order to include it in the journal.  

“The paper used a methodology that views the city as a network of connections between streets and uses that to make predictions about certain characteristics. At that time, it focused on gender violence and femicides in Mexico City”, said Palomares. 

“Many metrics for measuring the walkability of a public space don’t consider the human experience. I’m referring to safety, comfort, and thermal comfort”, he explained. 

He explained that AI is only used in architecture and urban planning to generate images. “The idea was to go beyond the image: how can we use AI to benefit city design and address issues like security?” he said.  

So, he furthered his research between August and November 2025. The paper now integrates several factors, including the human element. “I propose using hybrid intelligence that is not solely about AI”, he said. 

He complements it by including a computational agent for real data calculations and accurate information, plus a technique called graph machine learning and large language models (LLMs); “it’s seeing the city as a network”, he reflected.  

“Deep machine learning can help us make these predictions across the city, in street segments, and the human element is key to directing the whole process and making the important decisions”, he pointed out. 

Palomares tested his hybrid methodology to see if the predictions were more accurate, and the result was positive. His research has now evolved into the title “Hybrid urban intelligences: graph machine learning-driven multi-agent system for walkability". 

According to the academic, architecture has to some extent resisted integrating technology; his work aims to open doors to new ways of thinking.  

He recalled that he has spent two years delving deeper into AI, and above all, going beyond simply generating images. “I feel very excited and happy to share all of this”, he commented. 

His intention, he concluded, is to create tools that any urban planner or architect can use to develop or make predictions or diagnoses about cities.