Finalization of the 2nd phase of the modeling of the transport in the inner ring of barcelona
May 23, 2019
A new manufacturing system for integrated optical components
May 26, 2019The device, which is entering the pilot phase, is designed to assist families or carers of elderly people who live alone. It can detect hazardous situations such as whether the gas has been left on, the ventilation is insufficient, or food has gone off in the dwelling.
The UPC Research Centre for Biomedical Engineering (CREB) and the company Sensing & Control have launched a pilot test of a new system of artificial intelligence-based gas sensors to monitor elderly people who live alone.
The information gathered by the sensors is processed using algorithms developed by CREB and Sensing & Control’s applied artificial intelligence. The sensors have two main functions: they warn of hazards and provide information on the person’s pattern of behaviour. For example, whether a person has lunch or supper at irregular times, wakes up during the night or sleeps longer than usual could be indicators that “through artificial intelligence can be used to identify if an elderly person deviates significantly from their routine, which could be a warning sign”, explains the CEO of Sensing & Control, Narcís Avellana.
The system consists of two or three devices that are installed in the dining room, the bedroom and the kitchen of the dwelling. The sensors in the devices can detect various types of gases, such as carbon dioxide or hydrocarbons, among others, which reveal whether the person is in the space or not, has left the gas on, or whether there is insufficient ventilation or food that has gone off in the flat.
According to Jordi Fonollosa, a CREB UPC researcher, conventional monitoring systems, such as video cameras or movement sensors, “are often quite intrusive and create mistrust, compromise privacy and have limitations such as blind spots”, while gas sensors “can detect any activity that generates volatile substances; when we breathe, consume oxygen and release carbon dioxide, which result in a change in the air composition”. This solution is “more efficient because with fewer systems a larger area can be covered”, states Fonollosa.
For this innovation project, CREB UPC, a centre with the TECNIO seal, and Sensing & Control received a grant of 125,000 euros as part ACCIÓ’s INNOTEC framework programme (LLAVOR grant).
Sensing & Control (Barcelona) has seventeen employees and is a software SME specialised in the internet of things (IoT), artificial intelligence and smart homes.
Related Projects
- A research team from the Institut de Robòtica i Informàtica Industrial (IRI, CSIC-UPC), together with the Centre de Disseny d'Equips Industrials (CDEI) of the Universitat Politècnica de Catalunya - BarcelonaTech (UPC), are participating in the national CASANDRA project to develop a digital manufacturing solution that will make it possible to establish a bidirectional data flow for continuous digital integration throughout the supply chain across the product life cycle. This integration will be achieved through digital twins and data-based models that are continuously updated thanks to distributed monitoring and control tools.
- The research groups Structural and Materials Technology (ATEM) and the Construction Engineering (EC) at the Universitat Politècnica de Catalunya - BarcelonaTech (UPC) are working on the RE-STEEL project, which aims to develop an efficient methodology for assessing the rehabilitation of existing steel structures.
- The AgroTech research group at UPC, in collaboration with its spin-off Ugiat Technologies, has developed uPlayer, a new multimedia player concept that enables more intuitive video navigation and viewing, intelligently enhancing the user experience, especially on YouTube and other platforms, by integrating as a plugin or advanced player.
- The AgroTech research group at the Universitat Politècnica de Catalunya – BarcelonaTech (UPC), together with its spin-off Ugiat Technologies, have driven DoblAI, an AI platform that integrates transcription, translation, subtitling and video dubbing into a single workflow. The solution, which uses deep learning technology and cloned or default voice models, is specifically designed for the journalism and communications sector.
