As Qatar initiates more efforts to support people with autism, a research team from Qatar University (QU) has developed a device enabled with artificial intelligence (AI) to detect challenging behaviours in children with autism spectrum disorder.
Named, ‘SafeSignals’ and developed by a team led by John-John Cabibihan, professor of mechanical engineering and principal investigator of the project, the device relies on AI and machine learning technologies to analyse physiological and environmental indicators associated with stress and challenging behaviours.
Other members of the team are: Ahmad Yaser Alhaddad, Ahmad Qadeib Alban Malek Ayesh Abdulazizm Khalid al-Ali, Kishorn Kumar Sadasivuni and Hussein Aly from QU. The device was first showcased on the occasion of World Autism Day this year.
“We developed SafeSignals system to monitor and analyse physiological signals among children on the autism spectrum, through heart rate and perspiration levels. This innovation aims to empower parents, teachers, and caregivers by providing early alerts about a child's stress levels, enabling timely intervention when the child experiences sensory overload," said Cabibihan.
Through a video, the professor noted that sometimes the first sign of distress is not a shout or a scream. “It can be a signal under the skin. Children with autism often face sudden meltdowns from sensory overload. These are not ways of misbehaviour. It is the way of their bodies asking for help. Our team pairs wearable biosensors with social robots that sense rising stress in real time through heart rate, sweat response and movement which helps healthcare givers to act before a meltdown starts,” explained, Cabibihan.
He also narrated how in pilot sessions when the children's stress spiked, the robot gently reengaged them to calm down helping the movement safely and quietly. “These robots build bridges back to human connection because when technology listens to the body, you can step in early with care,” he added.
The SafeSignals system is a wearable and environmental sensor-integrated system. By combining machine learning algorithms, multimodal monitoring, and real-time alerts, it empowers caregivers and healthcare professionals to provide timely support, enhancing safety and wellbeing.
The main features of the device are multimodal monitoring as it integrates wearable and environmental sensors for holistic insights. It accurately identifies stress and challenging behaviours and provides real-time alerts by notifying caregivers and external devices instantly. It is reliable and scalable as it is designed to overcome limitations of current solutions with comprehensive monitoring system.
Moreover, the wearable sensors track physiological indicators of stress and anxiety and the environmental sensors monitor surroundings for potential triggers. With AI analysis, machine learning algorithms detect behavioural patterns and enables real-time intervention through immediate alerts to enable timely caregiver support.
The researchers believe that the device offers value proposition as it empowers care with timely, actionable insights for caregivers. It is future-ready and scalable for diverse healthcare and assistive applications.
The development of SafeSignals is part of collaborative research efforts by QU researchers to harness technology and innovation in advancing assistive technologies and strengthening healthcare-related fields. The team has already secured patent and is now seeking commercialisation opportunities.
Through partnerships with healthcare providers, assistive technology companies and investors, QU aims to expand the use of SafeSignals globally. With patent protection in place, the system is ready for licensing, co-development and integration into healthcare services.
