Ir a contenido principal
Finalista Nacional

NOBJECTE

NOBJECTE rethinks how blind people perceive and explore their environment with the objects around them in a way that is relevant. Co-designed together with Adrià, a visually impaired individual.

  • Meet Adrià, a visually impaired individual who collaborates with NOBJECTE to co-design the wearable.

  • A video introducing Adrià Esteve that captures the entire experience between him and NOBJECTE.

    A video introducing Adrià Esteve that captures the entire experience between him and NOBJECTE.

  • NOBJECTE, an AI wearable that helps to reduce the effort of a visually impaired person.

  • The User Interface of NOBJECTE, with two motor vibrators being activated in response to the trees.

  • The diagram provides an explanation of the technologies within the Raspberry Pi and its functioning.

  • NOBJECTE technology provides positive and haptic feedback to the user whenever an object is detected

Para qué sirve

NOBJECTE was created as a tool for blind individuals. It serves as to be able to alert and control their natural environment with the detection of objects. It helps them to reduce their effort every time they go out on the street without anyone's assistance.


Qué te inspiró

Adrià Esteve, a person who in 2020, at only 23 years of age lost his sight due to an eye disease. The advantages and disadvantages of technologies today is that they progress swiftly. But they don't always progress in the same direction such as for blind people. Adrià, like many other blind people, find it very easy to get disoriented on the street, either to take the same routes or to explore new places. With the AI technology that currently exists, we could update their tools to improve their daily life. In addition to proposing new attempts to design interfaces with AI outside of the most established ones such as the keyboard-screen.


Cómo funciona

NOBJECTE is a smart wearable that combines affordability and scalability co-designed with a blind individual, incorporating an image recognition machine learning model with individualised haptic feedback. The co-designed wearable works through a camera that visualizes the environment. Through Artificial Intelligence (the image recognition machine learning model called YOLO) connected to a Raspberry Pi, it possesses the ability of detecting objects through a Dataset*. Once the AI has detected the object in question, it sends a haptic “or a vibration” signal through the 5 vibration motors (coin motor vibrator) placed on the device and thus be able to locate and position the detected object in space. *The Dataset built into the AI is the main element in which NOBJECTE is going to detect objects. The Dataset can be easily exchanged by the user in order to tailor the experience and decide which objects to detect. NOBJECTE is Open Source and works with Python code.


Proceso de diseño

The co-design sessions with Adrià Esteve ended up with the iteration of up to 4 stages. The 1st was characterized by having all the technology located in the head, such as the battery, Rasperry Pi, vibration motors and camera. The camera was a 360-degree camera to have control of Adrià's environment. In the 4th iteration, we placed all the technology on the chest. As we were able to co-design the wearable unnoticed to everyone. To continue individualizing the experience of NOBJECTE with Adrià, we created a haptic language based on the two variables that vibration motors offer us. The power and the rhythm. We established all the requirements for him whenever an object was far or near. Finally, when we had the 4th iteration developed. We established two ranges of possibilities that were between an average situation and a difficult one to analyze the true value. In the average situation, the prototype was validated from Adrià’s house to Adria’s workplace. There we could see how Adrià's attention decreased, and he could concentrate on other aspects when he walked freely without anyone's help. During the difficult situation, it became a challenge that we would like to achieve in the future since we tested the device in a very large natural environment without reference elements.


Qué lo hace diferente

The different alternatives of our solution are guide dogs, which have a price of $50,000 in the USA and require a personal training of 2 years. NOBJECTE offers a fully customizable device tailored to the user's needs. In addition to providing a tool at a low price due to being Open Source, NOBJECTE can become a substitute for guide dogs in developing countries where guide dogs are only accessible to very few. NOBJECTE incorporates technology from artificial intelligence such as an image recognition models machine learning. It has the capability to choose which object to focus on thanks to the use of easily downloadable datasets, without the need for time-consuming investments. Furthermore, it provides blind individuals with the possibility to move freely and explore new environments.


Planes para el futuro

The possible prize acquired with this contest would provide the opportunity to acquire sufficient capital to standardize the object itself.Thanks to this, it would be possible to innovate technologically the tools used, such as smaller cameras and Raspberry Pi like the Arduino Nicla Vision that would go even more unnoticed with their use, providing increased durability and reliability when used. With an overall technological improvement, we will seek funding that will allow us to validate the product globally and obtain certifications to start production.In order to be able to give a positive response to Adrià Esteve and thus help my friend.


Premios

The project "NOBJECTE, an investigation on the exploration of new forms of interfaces design with Artificial Intelligence for blind individuals" has been nominated for the ELISAVA EI! Awards in the Technology category. ELISAVA EI are recognized as one of the top final degree projects of 2023.


Fin del contenido principal. Regresar al inicio.

Selecciona tu ubicación