top of page

Designing a Neural Input Wristband for XR Experiences - Sensor Fusion

In our previous blog post, we covered essential product specifications for neural input wristbands, discussing the ideal sensor count and wristband design. Now, let's delve into an exciting topic: sensor fusion. In this post, we'll delve into the concept of sensor fusion, where we bring together various sensor types to unlock an exceptional experience with the neural input wristband.


Sensor fusion is the art of harmonizing various sensor types to achieve a unified and accurate understanding of movements. When it comes to neural input wristbands, this fusion enables sensors to cooperate with finger and wrist motions, precisely tracking hand positions relative to the body. By blending different sensors, we overcome individual limitations and enhance interpretation reliability, ensuring smooth operation across diverse scenarios.


Sensor fusion's impact resonates with the evolution of technology. Just as multi-capacitive capabilities revolutionized input methods, gesture recognition sensors are spearheading the next phase of spatial navigation, particularly in smart wearables. This dynamic integration of sensors and signal conditioning units, forming transducers, underpins sensor fusion. These transducers convert physical actions into digital signals, offering new dimensions in usability and interaction. As we stride into the future, sensor fusion promises to reshape our relationship with technology, delivering seamless, intuitive, and immersive experiences.

The relevant discussion begins at 36:54


Internal sensors measure electrical signals such as resistance, capacity, photonics, ionization, and magnetism. These signals can originate in the user as bio-potentials, or in the user surroundings. Internal sensor examples are EMG, EEG, PPG, and SNC. Because all sensor types have limitations, sensor fusion improves the probability that there will always be a sensor that performs well in any given situation. Each sensor must work independently without any interaction between them. Thus, correct decisions always will be made, regardless of user physiology or scenario conditions that are challenging for a certain sensor type.

Hand held, Camera, and SNC interface performance across different dimensions

Source: Wearable Devices Ltd.


Keyboard and Mouse: The keyboard, a static input device, uses external sensors to detect key presses, suitable for typing and numeric data entry. The mouse, a dynamic input device, relies on external sensors for movement and clicking. It offers precise cursor control but requires a flat surface and has limited gesture functionality. Both demand a static surface, hindering mobility and adaptability.


Touchpad and Touchscreen: Touchpads, dynamic input devices, detect finger movement and gestures, used in laptops and portable devices. They're intuitive for cursor control but less precise than a mouse. Touchscreens enable direct interaction but lack tactile feedback, susceptible to accidental taps and visibility issues. Extended usage might cause physical strain.


Gaming Controller: Gaming controllers use a mix of external and internal sensors for buttons, joysticks, and orientation. Suitable for gaming, they offer diverse input options but can be complex and require learning.


Gesture Camera: Gesture cameras employ internal sensors to track hand and body movements. They allow hands-free control and offer a natural interface but require a clear view, are affected by conditions, and demand specific gestures.

Voice: Voice activation is hands-free and eyes-free, useful when occupied or needing visual attention. Ambient noise and accents can affect performance, requiring specific commands and drawing attention in public.


Bio-potential sensors, such as EEG, EMG, PPG, and SNC sensors, measure electrical activity in the brain or muscles, blood volume changes, and pressure gradations, offering natural and non-invasive interfaces but require expertise to interpret the signals and can be affected by external noise and factors.


Surface Nerve Conductance (SNC) sensors are a type of non-invasive sensor used in wearables. They react to ions and detect innervation mostly through wrist movements. These sensors offer a non-invasive and convenient way to track physiological signals, such as real-time measurement of pressure gradations, and they can be highly accurate in detecting minute changes in the body. However, although these effects can be mitigated, SNC sensors can be affected by various sources of noise such as humidity and sweat, which can decrease their accuracy. Additionally, calibration may be necessary to ensure accurate measurements, and these sensors require constant physical contact with the skin.


We would like to exploit the strengths of such sensors and therefore we need sensor fusion to extract all the information available for the highest possible accuracy and reach.


In our upcoming and concluding post of this series, we unveil a compelling argument for the indispensable role of wearables that are touchless and hands-free within the realm of extended reality (XR). We delve into why these attributes stand as XR's minimum viable product (MVP). Additionally, we proudly introduce the Mudra Band – an innovative aftermarket wearable input wristband that embodies our vision and insights.


*All figures shown in this blog are taken from our white paper, available for download here.


Comments


bottom of page