Saturday, September 20, 2025

 

PolyU scholars pioneer smart and sustainable personal cooling technologies to address global extreme heat




The Hong Kong Polytechnic University
PolyU scholars pioneer smart and sustainable personal cooling technologies to address global extreme heat 

image: 

Prof. Dahua Shou, Limin Endowed Young Scholar in Advanced Textiles Technologies, Associate Professor of the PolyU School of Fashion and Textiles, Associate Director of the Research Centre of Textiles for Future Fashion, and Associate Director of the PolyU–Xingguo Technology and Innovation Research Institute, has published a peer-reviewed paper in Science, offering new insights into sustainable personal cooling using advanced textiles and intelligent wearables.

view more 

Credit: © 2025 Research and Innovation Office, The Hong Kong Polytechnic University. All Rights Reserved.






Global warming poses a growing threat to human health and work performance. Currently, about 3.6 billion people worldwide live in areas highly susceptible to climate change. From 2000 to 2019, more than 480,000 heat-related deaths occurred globally each year. Extreme heat also impairs focus and productivity and worsens mood by elevating stress hormones and disrupting sleep. In response to the increasing frequency of heat waves, The Hong Kong Polytechnic University (PolyU) scholars are developing next-generation personal cooling solutions that push the limits of conventional clothing and promote sustainability.

Prof. Dahua SHOU, Limin Endowed Young Scholar in Advanced Textiles Technologies, Associate Professor of the PolyU School of Fashion and Textiles, Associate Director of the Research Centre of Textiles for Future Fashion, and Associate Director of the PolyU-Xingguo Technology and Innovation Research Institute, has published a peer-reviewed paper in Science, offering new insights into sustainable personal cooling using advanced textiles and intelligent wearables.

Smart technologies, especially intelligent wearables and AI, can be key to sustainable personal cooling. Prof. Dahua Shou said, “According to the World Meteorological Organisation, there is an 80% chance that at least one year between 2025 and 2029 will be the hottest on record, making personal cooling increasingly vital for well‑being, health and productivity. We have been creating intelligent, superhero‑like garments that provide on‑demand adaptive cooling and clinician‑like health monitoring to help address the challenge of extreme heat.”

By integrating the four cooling mechanisms of radiation, conduction, convection and evaporation, this stand‑alone perspective outlines strategies to adaptively regulate body heat and moisture in dynamically changing real‑world settings. The paper also presents an AI‑driven, closed‑loop framework that connects sensing, prediction, and actuation to deliver personalised, energy-efficient cooling, with a scalable and recyclable design that supports public health, workplace safety, and performance.

Sustainable personal cooling is evolving from the use of passive fabrics to the integration of smart systems. Notably, spectrum‑selective textiles effectively release mid‑infrared body heat while blocking external solar and urban heat gain. Thermal insulation is being engineered with conduction-tunable fillers, while ventilative and evaporative cooling is boosted by moisture-responsive fibres. Lightweight wearables, such as variable emittance devices, and electrocaloric and thermoelectric modules paired with flexible solar and on-body energy storage, enable active and controllable cooling. These emerging technologies strategically employ model-selective cooling and incorporate human‑centered design for comfort, durability, washability, and low weight, expanding comfort zones and reducing dependence on air conditioning.

Despite promising progress, key challenges in personal cooling remain. Sweating helps cool the body, but limited sweat management increases fabric weight and cling, while reducing permeability and radiative cooling efficiency, especially during heavy perspiration. Real-time adaptive thermoregulation, which responds to changing environments and individual physiology while ensuring comfort and safety, is ideal but difficult to achieve.

Prof. Dahua Shou said, “We also need interdisciplinary integration across textiles, thermodynamics, flexible electronics, and AI, along with scalable, recyclable manufacturing that balances sustainability, wearability, fashion, and performance. Standardised, user‑centric metrics, such as cooling power per watt, thermal sensation, and user acceptance, are essential for fair comparison and adoption.”

Prof. Shou and his research team are tackling extreme heat with various innovative technologies. iActive™ intelligent sportswear uses low‑voltage-driven artificial “sweat glands” and a root‑like liquid network mapped to sweat zones to quickly eject perspiration as droplets, reducing weight and cling, keeping the skin dry, and removing sweat up to three times faster than peak human sweating.

Omni‑Cool‑Dry™ is a breathable skin‑like fabric that routes sweat directionally while providing spectrum‑selective cooling. By reflecting solar and ground radiation and emitting mid‑infrared body heat, it helps keep wearers cool and dry even under the sun, lowering skin temperature by about 5°C compared to conventional fabrics.

For hot workplaces, thermo‑adaptive Soft Robotic Clothing embeds temperature‑responsive soft actuators that expand to thicken fabric and trap still air, solving the problem of “one‑level” thermal insulation. Thermal resistance varies from 0.23 to 0.48 K·m²/W, keeping the inner surface 10°C cooler than conventional insulating garments even when the exterior temperature reaches 120°C.

SweatMD is an all‑textile, non-invasive wearable that channels fresh sweat through a nature‑inspired microfluidic network and uses skin‑friendly sensing yarns to track biomarkers like glucose and potassium. It delivers real‑time, molecular‑level health insights such as indicators of fatigue and dehydration to a smartphone.

Collectively, these innovations form an AI‑ready ecosystem: sensors quantify physiology, models predict cooling demand, and intelligent clothing actuates targeted responses. Integrating textile sensors, fiber‑based coolers, and on‑body energy harvesters has the potential to enable self‑sustained cooling.

Spanning everyday wear, as well as sports, and protective gear, these innovations bridge the gap between fundamental research and real-world applications to address global challenges. PolyU translational research institutes across Mainland cities and interdisciplinary research centres, such as the PolyU-Xingguo Technology and Innovation Research Institute and the Research Centre of Textiles for Future Fashion, allows the University’s scholars to tap into these cities’ diverse application scenarios and collaborate with leading enterprises to accelerate the transformation and scalable deployment of scientific research achievements.

These research innovations earned the Gold Medal with Congratulations of the Jury (2025) and a Gold Medal (2024) at the Geneva Invention Exhibition, as well as the TechConnect Global Innovation Award. Prof. Shou also received The Fiber Society’s Distinguished Achievement Award, a prestigious honour awarded annually to a single scholar worldwide.

New digital tool helps police forces identify potential victims of human trafficking  




University of Sheffield



  • The Sexual Trafficking Identification Matrix (STIM) tool developed by the University of Sheffield is helping law enforcement, charities and online platforms in their work to identify and support potential victims of sexual exploitation and human trafficking

  • Previously organisations would potentially need to look through thousands of adverts online manually everyday, which can take a massive toll on those doing the work

  • The new tool has already resulted in multiple arrests and five live criminal investigations by Thames Valley Police as well as helping to safeguard dozens of vulnerable people

  • STIM can be used to assess adverts posted on adult service websites for risks that suggest they are involved in human trafficking or sexual exploitation

  • The tool is already being used by charities and organisations world-wide and has been recognised as an example of good practice by UK policing

A new analytical tool that helps identify potential victims of human trafficking and sexual exploitation is helping police forces across the UK work to get justice for victims.

The Sexual Trafficking Identification Matrix (STIM) is used to assess adverts posted online for risks that suggest the people involved may be victims of human trafficking or sexual exploitation.

Developed by Dr Xavier L'Hoiry from the University of Sheffield, in collaboration with global partners, STIM is already proving to be vital for organisations like the Thames Valley Police, in its work to proactively identify and protect individuals at the greatest risk of exploitation.

A major obstacle to this work is distinguishing between profiles created by non-trafficked sex workers and those posted by traffickers, as it's difficult to tell a legitimate profile from a false one.

STIM works by assessing adult service website adverts by analysing elements like the photographs, language and details used to produce a rating to decide whether the advert is low, medium or high risk. This can then be used by practitioners to decide whether to dedicate additional resources for investigations, or support for potential victims.

Dr L’Hoiry, from the University of Sheffield’s School of Sociological Studies, Politics and International Relations, worked with various national and international partners to develop the tool, including the Universities of Northumbria and Copenhagen, law enforcement, charities and NGOs - including the The Danish Centre against Human Trafficking (CMM) which works to identify and support victims of human trafficking.

He said: “In the digital age, the use of some adult service websites by criminals presents a growing challenge in the battle against human trafficking and there is growing evidence that traffickers are using them to reach a larger customer base and operate with less risk of being caught by law enforcement. 

“By developing a tool like STIM, we wanted to help those working to tackle human trafficking to be more efficient in their efforts to reach out to those vulnerable to sexual exploitation. Previously organisations would need to look through thousands of adverts online manually, which can take a massive toll on those doing the work. The STIM still relies on human intervention but it aims to make the risk assessment process much quicker and straight-forward for those involved.”

STIM has so far been adopted by 16 police forces and two charities in the UK, four police forces abroad, and the Danish Centre Against Human Trafficking in Denmark (CMM). Thames Valley Police have been the force leading the way in its implementation in the UK.

Ben Henley, a Detective Chief Inspector who has been leading the STIM project with Thames Valley Police said: “One of the main challenges we had was that we were reliant on victims coming forward, which was unfortunately very rare. So the STIM has been really beneficial for us. It has helped us identify those at greatest risk and set timeframes on how quickly we action those.

“We’ve utilised the STIM to score over 128 online adverts, and this has resulted in 40 visits being conducted at locations, with dozens of individuals safeguarded. We currently have five live criminal investigations and there have been multiple arrests as a result of utilising the STIM. Other forces have made contact with me and I've provided briefings to them so they can replicate our model. It has also been identified as national best practice and as a result, other forces have been encouraged to follow our utilisation of the STIM.”

Dr L’Hoiry and colleagues from Northumbria and Copenhagen are continuing to update and improve on STIM based on the feedback of the partners using it and offering training and supporting its roll-out in police forces and other organisations, with so far 30 sessions delivered to over 100 police representatives.

He added: “STIM has already delivered many benefits to users,  taking our research beyond the walls of the University to connect with the organisations making a difference for potential victims. Our ambition is that every police force in the UK, and charities and organisations in the third sector, here and worldwide, adopt the use of the STIM tool to help assist their work in identifying victims of human trafficking and developing meaningful interventions to support them.”

 

ENDS

Dogs can extend word meanings to new objects based on function, not appearance



Dogs with a vocabulary of toy names—known as Gifted Word Learners—can extend learned labels to entirely new objects, not because the objects look similar, but because they are used in the same way.




Eötvös Loránd University

Dog brings the toy 

image: 

A new study publishing in Current Biology on September 18 reveals that dogs with a vocabulary of toy names—known as Gifted Word Learners—can extend learned labels to entirely new objects, not because the objects look similar, but because they are used in the same way.

view more 

Credit: Department of Ethology / Eötvös Loránd University





BUDAPEST, Hungary — A new study publishing in Current Biology on September 18 by the Department of Ethology at Eötvös Loránd University reveals that dogs with a vocabulary of toy names—known as Gifted Word Learners—can extend learned labels to entirely new objects, not because the objects look similar, but because they are used in the same way.

In humans, “label extension” is a cornerstone of early language development. In non-humans, until now, it had only been documented in few so-called language-trained individual animals, after years of intensive training in captivity.

But learning to extend labels to objects that share the same function, rather than visual similarities, is considered an even more complex skill. A toddler learns that the word “cup” can apply to mugs, tumblers, and sippy cups, or that both a spoon and a ladle are “for scooping.” While individuals of many animal species can group items by appearance, extending a learned label to a functionally similar but visually different object has long been considered an advanced skill.

Video abstract at this link: https://youtu.be/8_NbCYAWSfU

The time and efforts needed to train animals in captivity to learn verbal labels, as well as the very limited number of subjects that successfully acquired such vocabulary, have until now limited the feasibility of this type of research.

But here comes the twist! “Gifted Word Learner dogs offer a unique possibility to study this phenomenon because they rapidly learn verbal labels – the names of toys – during natural interactions in their human families” said Dr. Claudia Fugazza, lead author of the study.

“Our results show that these dogs do not just memorize object names,” continues Dr. Fugazza. “They understand the meaning behind those labels well enough to apply them to new, very different-looking toys— by recognizing what the toys were for.”

Link to the social media of the Gifted Word Learner dogs project: https://linktr.ee/geniusdogchallenge

A Play-Based Experiment

Researchers of the Department of Ethology, at Eötvös Lorand University tested 7 Gifted Word Learner dogs—(six Border collies and a Blue heeler)—known for their unusual ability to learn the names of dozens of toys naturally, through everyday play.

The experiment had four stages, all of them conducted in a natural setup, at the house of each dog owner, during playful interactions:

  1. Fist, in the Learning Phase, Dogs learned two new labels, such as “Pull” and “Fetch,” each referring not to a single item, but to a group of toys that looked completely different but were used in the same way during play (tug or retrieve).
  2. Second, during a formal Assessment, the dogs showed that they had successfully learned those labels and could appropriately choose the “Pulls” and “Fetches” when asked.
  3. The crucial part of the experiment was carried out after this Assessment: in the Generalization Phase, the dogs were introduced to new toys, also with diverse physical features, and the owner played in the same two ways as before, but this time saying no labels.
  4. Test – When asked for a “Pull” or “Fetch,” the dogs selected the correct unlabelled toy significantly above chance, indicating they had generalized the labels to a functional category.

Why This Matters

The study provides the first evidence that dogs can generalize verbal labels to functional categories during natural-like playful interactions in their human families—mirroring, in functional terms, the natural context of human language development.

“This ability shows that classification linked to verbal labels can emerge in non-human, non-linguistic species living in natural settings,” said Dr. Adam Miklosi, coauthor of the study. “It opens exciting new avenues for studying how language-related skills may evolve and function beyond our own species.”

Key Points

  • Dogs extended verbal labels to objects that shared only functional properties, not appearance.
  • The skill emerged naturally through play with owners—no formal training required.
  • While the mechanisms of such learning are not known, the context in which it happens present a striking parallel with that of human infants: daily life in a human family.
  • The study of these skills in a non-human species in its natural environment paves the way for understanding the how language-related skills evolved and function.