AI redefines fingerprint analysis

Representation Pic

Columbia University engineers have made a novel discovery in forensics with the development of a new AI system. Contrary to the long-held belief that fingerprints from different fingers of the same person are unique, their research reveals that these fingerprints are actually similar.

This finding emerged from analysing a public US government database of 60,000 fingerprints using an AI system. The system's ability to identify similarities in fingerprints from the same person improved over time, reaching an accuracy rate of 77 per cent for single pairs. The research was published in Science Advances after rejections from forensic journals.

New way to recycle fabric found

Researchers at Aarhus University have developed a new technology for recycling fabrics mixed with elastane, a material found in most sportswear and increasingly in everyday clothing. Elastane, known for its elasticity, has been a major hurdle in textile recycling due to the difficulty in separating it from other fibers like cotton, wool, and nylon. This has contributed to poor recycling rates.

The new method involves heating the clothes in a large "pressure cooker" at 225 degrees Celsius with a specific alcohol and potassium hydroxide. This process effectively breaks down the elastane fibers, allowing them to be separated from other materials. Though currently efficient with nylon, the team is adjusting the process to fully integrate cotton.