A Cognitive Assistant (CA) is a relatively new concept focused solely on direct assistance to people with cognitive or physical disabilities, expanding the area of Ambient Assisted Living (AAL) to include complex platforms that include sensors, actuators, monitoring capabilities and decision processes. CAs focus on individuals and their disabilities, providing the tools that best suit them through customization methods. CAs can range from a medication reminder to a complex mentoring system that connects its users with their family members. New developments such as the Internet of Things and Edge Computing have enabled the development of wellness-enhancing environments that were previously unavailable through embedded systems.

Today, physical cognitive assistants have evolved thanks to Edge AI-based platforms (combination of Edge Computing and Artificial Intelligence). Edge AI proposes that AI algorithms are processed locally on the device itself or on a server close to the device. AI algorithms use data obtained by the devices themselves at a low level and can make decisions quickly without having to transmit large amounts of information to the cloud. Examples of such physical cognitive assistants include: ElliQ elliq.com and Senter senter.io, which aim to support a healthier and happier aging population. One of the drawbacks of the assistants available on the market is their high price, which makes them unaffordable for a significant percentage of the population and which usually coincides with the population at risk of exclusion and more vulnerable. However, with the emergence of new, smaller, cheaper and more powerful devices, it is possible to adapt a new type of low-cost cognitive physical assistants that can be accessible to the widest possible range of the population.
Accordingly, GTIbot is the software associated with a cognitive physical assistant in the form of a companion robot that monitors certain biosignals for elderly people or people with specific disabilities. Specifically, the assistant attempts to determine the person’s state of health and mood using its built-in camera.

GTIbot has the ability to recommend the performance of exercises proposed by its caregivers. All of the person’s activities are monitored and stored for possible follow-up by the caregiver. Using the detection of emotional states, the assistant can determine the user’s emotion during the performance of the exercises, during their daily or scheduled chores. Another important aspect of the robot is the integration of environmental sensors. These sensors can measure levels of humidity, temperature, CO2, total volatile organic compounds (TVOC). The information provided by these sensors allows the robot to inform the caregivers if any of these values exceed levels that could threaten the integrity of the person.
The acquisition of heart signals (ECG) allows GTIbot to monitor elderly people with heart problems, the robot has the ability to communicate with a wearable device that capture ECG and send it to the robot using Bluetooth communication for analysis, allowing to determine if the user may have atrial fibrillation.