Pixels, ants and DNA: Big innovation awaits in small packages

In a world where cutting-edge technology often springs from the laboratories of Silicon Valley, some innovative solutions are being drawn from more unexpected sources. Researchers are taking cues from video games and ants and applying large language models in unusual ways to solve complex real-world problems.

Building better blocks with polygons and pixels

Picture a thriving metropolis with high-rise buildings, expansive parks and streamlined public transportation. Could this city's design have stemmed not only from from expert urban planners, but from passionate citizens who are also avid city-building game enthusiasts?

New research from Lancaster University in the U.K. aims to revolutionize urban planning by incorporating input from the city’s children into the design. Published in Acta Ludologica, a peer-reviewed scientific journal about the discourse of games and digital games, the study illustrates the lack of public engagement in current urban development practices and suggests using gaming platforms like a modified “Cities: Skylines” or “Sim City” to offer realistic simulations for residents, enhancing their involvement and awareness in city planning.

The technology at the heart of this study is a sophisticated modification of “Cities: Skylines,” which allows players to import real-world buildings and models to create lifelike urban environments. Participants can manage aspects of city life, including education, public services and tax policies, while the game dashboard tracks the happiness of citizens. This interactive approach not only educates players about the intricacies of urban planning but also serves as a tool for real-world applications. Researchers Paul Cureton and Paul Coulton, from Lancaster’s design-led research lab ImaginationLancaster, have demonstrated the effectiveness of this method through workshops with the Lancaster City Council after engaging children in the planning of a new garden village.

The implications of this research are profound. By integrating game design with urban planning, the researchers offer a cost-effective, enjoyable and scalable method for increasing citizen engagement in the planning process. This approach also addresses the urgent need for change in public participation, as highlighted by the Royal Town Planning Institute's data indicating minimal interest in planning among younger people.

The study concludes  by suggesting that such innovative use of gaming technology could support planners, enhance skill development and provide the necessary tools to involve people more deeply in the transformation of their living spaces. Ultimately, this research paves the way for a more collaborative and dynamic future in urban development.

The language of life

The next stop on our look at unconventional tech inspirations takes us to the microscopic world of DNA. Researchers at Germany’s TU Dresden are using artificial intelligence to unlock the hidden language of DNA, providing new insights into genetics and disease.

DNA is often described as the blueprint of life, containing all the instructions for building and maintaining an organism. However, deciphering all of the information within DNA is incredibly complex and not yet fully understood, and traditional methods of DNA analysis can be slow and labor-intensive. That’s where artificial intelligence comes in.

Researchers at TU Dresden have developed a new AI model named GROVER (Genome Rules Obtained via Extracted Representations), which treats DNA sequences like a language, using techniques like those employed in natural language processing. By analyzing patterns and structures within the DNA code, the AI can identify sequences. This method, akin to deciphering a foreign language, allows for researchers to employ GROVER for faster and more accurate interpretation of genetic data.

By training on the entire human genome, GROVER creates a DNA dictionary that researchers hope can unlock insights into genetic codes, advancing genomics and personalized medicine. This research, published in Nature Machine Intelligence, holds potential for significant breakthroughs in understanding DNA’s complexity.

“In terms of language, we are talking about grammar, syntax and semantics," Melissa Sanabria, a researcher behind the project, told the university’s website. "For DNA, this means learning the rules governing the sequences, the order of the nucleotides and sequences, and the meaning of the sequences. Like GPT models learning human languages, GROVER has basically learned how to ‘speak’ DNA.”

Analyzing ants to mobilize miniature machines

Researchers at the TU Delft MAVLab have made a significant breakthrough in robot navigation by drawing inspiration from how ants navigate their environment. Ants use a combination of visual recognition (think snapshots) and step counting (think odometry) to return home, even after long journeys. The MAVLab has emulated this method, creating an insect-inspired navigation strategy for tiny, lightweight robots.

Traditional autonomous navigation systems often rely on hardware that can be impractical for tiny robots. Inspired by nature, MAVLabs researchers have designed a system where robots take snapshots of their surroundings to aid navigation. This method, akin to Hansel and Gretel’s breadcrumb trail, involves the robot taking visual snapshots at intervals and using them to guide its return journey. By combining these snapshots with odometry,  the robots cover greater distances more efficiently, significantly reducing the computational burden.

This bio-inspired approach has been successfully demonstrated with the small drone called  the CrazyFlie, which can navigate up to 100 meters using only 1.16 KB of memory.  The research paves the way for the practical deployment of tiny robots in various real-world scenarios where traditional navigation methods may be unfeasible. While the system doesn’t generate detailed maps, it provides a robust solution for specific tasks like inventory tracking and greenhouse crop monitoring, where simple return-to base functionality is sufficient.

Whether it’s using video games to improve urban planning, drawing on the navigation skills of ants to enhance robotics or employing AI to decode the language of DNA, these progressive examples demonstrate the power of thinking outside the box to drive remarkable advancements. And as researchers continue to look at the natural world and beyond for inspiration, we can expect more of these innovative solutions to emerge, transforming the way we live and interact with technology.

Banner image of a city simulation was developed with generative AI. Photo credit: Adobe Stock