Assistive technology has come a long way over the past decade in allowing people who couldn’t ordinarily use a mouse and keyboard to still interact with a PC or gaming console. However, their often high cost put them out of the reach of many, which is why Alex Dunn created a lineup of AI-enabled controllers that are powered by boards from the Raspberry Pi Foundation.
The goal of Enabled Play is to introduce new ways to control computers through methods such as voice commands, facial expressions, body gestures, and virtual buttons in order to make gaming more accessible. Each device communicates over Bluetooth to a host PC in order to send data. Interpreting human movements into keyboard and mouse inputs is accomplished by taking advantage of edge machine learning, wherein a model runs locally on the computer and produces a corresponding output, such as recognizing speech or gestures. An accompanying mobile app can then adjust certain settings and change what human inputs map to a given action.
The Enabled Play Controller is the first in Dunn’s lineup of assistive devices. It supports inputs ranging from offline speech, facial, and gesture recognition through the use of built-in microphones and an optional camera, motion controls for tilting and shaking, and virtual buttons. With these possible methods of gathering data, there are countless combinations of personalized experiences that can be created for this highly accessible human input device.
In addition to the primary Controller, Dunn also came up with a smaller Lite variant that can achieve very similar goals to the normal version, albeit with some lesser performance due to it being based around a Raspberry Pi Zero W rather than a Raspberry Pi 4 Model B in the former. Both controllers even have support for automatic updates so that new features can be added over time.
Currently, Dunn is working on a second version of both controllers that contain substantial upgrades over the previous models. The Enabled Play Controller v2 houses a Raspberry Pi Compute Model 4 mounted onto a custom carrier board for even better portability while still maintaining its great feature set. The Lite model’s Raspberry Pi Zero W was swapped for the newer Raspberry Pi Zero 2 W, although it no longer includes support for computer vision.
Getting started with the Enabled Play ecosystem is extremely easy due to the impressive accompanying mobile app. Users simply register for an account, add their new Controller, and can start sending it commands immediately. From here, the app allows people to create profiles that store inputs and the resulting keys, buttons, and/or mouse movements to make in response. For more information about the Enabled Play Controller lineup, you can visit its website here.