Creating a system powered by artificial intelligence that improves the mobility of the blind by 25%

 Creating a system powered by artificial intelligence that improves the mobility of the blind by 25%

Autonomous mobility is a huge challenge that blind people face daily, and for decades the white stick has been the main tool of assistance, but its limited scope and inability to determine the nature of obstacles stand in the way of smooth and safe mobility, especially in complex environments. In a promising step towards overcoming these challenges, a team of Chinese researchers has developed an innovative wearable system based on integrating artificial intelligence into glasses equipped with a camera and earphones to provide accurate mobility guidance for the blind and visually impaired, offering an alternative that surpasses its advantages over traditional means.

But what is the mechanism of action of this system

Creating an AI-powered system that improves the mobility of the blind by up to 25%

This innovative advanced system relies on integrating several technologies together to achieve its goal, as the main part consists of a pair of glasses equipped with a camera that takes direct and continuous images of the environment surrounding the user and then sends these images to a compact microcomputer that processes them instantaneously using advanced machine learning algorithms. These algorithms have been trained accurately to identify various elements in the user's surroundings, such as other people, doors, walls, pieces of furniture, and other possible obstacles.

Based on the analysis of the images, the device emits sound alerts to the user every 250 milliseconds, which provides an almost instant update of spatial information, and these alerts come in the form of a characteristic beeping sound in the right or left earphone to guide the user towards the right direction to avoid a collision.

To enhance the accuracy of the navigation aid, especially at close distances, the researchers created a unique additional component, which is elastic patches resembling artificial skin worn by the user on his wrists and fingers. These patches come in the form of bracelets equipped with a special camera system to act as additional tactile sensors. They vibrate to alert the user when he approaches an obstacle within a distance of 5 centimetres to 40 centimetres, providing a vital additional layer of protection.

These patches also perform another very important function, as they vibrate when the user reaches out to hold an object, informing him of the right moment to hold it, making it easier to interact with surrounding objects.

Promising experiments and promising results:

To assess the effectiveness of the system, the researchers, according to a study published in the prestigious journal Nature Machine Intelligence, conducted a series of experiments with the participation of people with visual impairments. In one of the main experiments, the team recruited twenty participants to test the ability of the device to help navigate inside a closed maze 25 metres long.

The results showed a significant improvement; the participants ' walking distance increased, and their commuting time improved by up to 25% compared to their performance when using a traditional white cane.

In other experiments involving 12 visually impaired people, the researchers aimed to assess the ability of the system to help avoid obstacles while walking inside rooms, and after undergoing simple training on using the device, all participants were able to successfully navigate inside buildings at a walking speed similar to their speed when using a stick, indicating ease of learning and adaptation to the system.

The tests were not limited to controlled environments; they extended to more complex real-life situations, as 8 participants tested the device in scenarios from everyday life, including walking through busy city streets and passing meeting rooms full of furniture and various obstacles, and in these difficult conditions, the system continued to perform well, confirming its ability to deal with diverse and dynamic environments.

Advantages and future challenges:

As Dr. Leilei Gu, a researcher in the field of artificial intelligence at Shanghai Jiao Tong University in China and one of the participants in the study, points out, the cameras have a wide viewing range that exceeds the one metre that the stick is usually able to sense. More importantly, the system not only touches the environment as a stick does but also has the ability to find out what the surrounding objects are thanks to artificial intelligence, which gives the user a deeper understanding of his surroundings.

For his part, Dr. Botund Ruska, director of the Institute of Molecular and Clinical Ophthalmology in Basel, Switzerland, gave a preliminary assessment of this development, saying, “The research paper deals with how to develop a super-intelligent stick. In this way, he likens the system to a stick in terms of being a tool to expand the user's sensory perception and help him navigate, but recognises that this system provides much richer and more complex information.

Therefore, Dr. Ruska believes that this smart system may be more useful and effective, especially for people living in large and crowded cities, as the surrounding environment requires greater awareness and faster reaction to changing obstacles.

But at the same time, Dr. Leilei Gu emphasises that the system in its current state is still just a prototype undergoing testing and development. He said, “In order for this system to become a practical and useful tool that people with visual impairments can rely on in everyday life, the research team must verify that it reaches super and very high levels of reliability to ensure the accuracy of information and directions, in addition to achieving maximum safety for the user, to avoid any potential risks that may result from any error in estimating distances or identifying obstacles.

Dr. Ruska agrees with this opinion, saying, “It is too early to say how much users will actually accept this technology when it becomes available.” He pointed out that the acceptance of such a technology will depend on multiple factors beyond just the technical effectiveness in the laboratory, including the final cost of the device, ease of use and learning, the convenience of wearing glasses and other components for a long time, and actual performance in a variety of unpredictable outdoor environments.

Therefore, the research team is currently working on the development of the device, focusing on improving the design to make it lighter and less visible compared to the current model in which the cameras are mounted on the glasses, to be comfortable and easy to wear daily.

This AI-based system represents an important step towards enhancing the autonomy and freedom of movement of people with visual impairments. By providing real-time and detailed information about the surrounding environment and accurate alerts, this technology promises to improve the quality of life and increase safety for its users.