In the previous blog post we delved further into the world of HCI (Human-Computer Interaction) and BCI (Brain-Computer Interfaces), and presented out new taxonomy that defines the levels of interaction. Levels 0-2 focused on touch and muscle based interactions while levels 3-5 use intent and neural-centered command inputs. In this blog post, we will explore the tradeoffs and balance between functionality, accuracy and design when creating a neural input device while keeping the user at the forefront.
When developing a wearable neural interface, there is always a tradeoff between functionality, accuracy, and design:
Functionality – determining the scope of features and capabilities and the input types it offers
Accuracy – an input method that is reliable and provides high accuracy of the intended functionality for all types of user physiology
Design – a wearable interface that is comfortable, durable, stylish, and fits a user's daily routine
Creating a successful consumer-level neural interface product involves balancing numerous factors like weight, size, materials, and user experience. It's crucial to consider components like sensors that detect neural signals, algorithms that interpret data, and comfortable hardware. Choosing, designing, and integrating these elements carefully is vital for an effective and user-friendly device. Success hinges on accurately detecting neural activity and providing valuable feedback to enhance the user's experience.
The relevant discussion begins at 16:53
A neural input device specification needs to focus on three main aspects: the user, the wearable product, and the interaction between the user and the product. The technologies involved can be grouped into three levels: hardware, software, and humanware, which involves incorporating a human perspective into the development of both hardware and software. These categories have six layers, and each layer relies a lot on the ones next to it. To create similar solutions, you need to excel at each of these layers.
Hardware includes band design and material, sensors and electrodes which make up the interfacing hardware and the electronics including a miniaturized flex-rigid dynamic PCB.
Software includes AI learning algorithms. The artificial intelligence algorithm is tuned to run on devices with limited power and processing resources.
Humanware is the hand and finger movements,the functions binding these interactions and the user’s performance.
Humanware should be placed first. Lasting success will come from a natural, intuitive, user-friendly and comfortable product. The hardware and software should prioritize a user-friendly and intuitive experience, considering how humans naturally interact. Designing should begin by focusing on the user's experience, which helps create a functional and easy-to-use product. Things like the device's size, shape, materials, and how users interact with it must be thought about. This approach ensures the device meets user needs and goes beyond, providing comfort, usability, and satisfaction.
Further factors to take into account include:
The device's location on the body - The electrodes on the wristband are placed on the inside of the wrist. This spot makes it easy to capture and understand neural signals, which can then be used to control different things. It's also simple to connect it to a smartwatch, and it doesn't disrupt any sensors already on the watch's back.
The device needs to be fashionable - a wristband form-factor is both functional and fashionable. It supports multiple stylish designs to make it an attractive accessory.
The device needs to be comfortable - the wristband is made of soft and flexible bio-compatible materials that ensure a comfortable fit for extended all-day wear.
Surface Device vs. Implant - there is a tradeoff between the two. A neural input wearable is like a surface-level gadget, which means you don't need surgery to use it. This makes it safer and easier for lots of people to use. Wearables are less intrusive and usually cost less than implants, but they might not be as accurate in reading neural signals and could be affected by things around them. Implants are super accurate at reading neural signals. They aren't influenced by outside things and give a more solid neural input. But, getting an implant requires surgery, costs a lot more, and could have problems like infections, rejection, or harm to nearby tissue.
In the next blog we will define the specification guidelines for developing a neural input wristband and explore system specification parameters such as reach, accuracy, latency, computing power, and product specifications such as the number of electrodes and the band dimensions.
*All figures shown in this blog are taken from our white paper, available for download here.