I'm looking for a standalone voice assistant designed around accessibility rather than cloud services. Ideally, it would work without sending data to the internet, would not require a phone to serve as the main device, and would provide a synthesized voice for reminders, alarms, routines, calendar events, and custom spoken messages. I'd also like local volume controls, adjustable speech speed, and the ability to choose a custom wake word or name.
A useful system might let me create commands such as, "Every Tuesday at 6 p.m., remind me to take out the trash by saying, 'The trash needs to go out tonight.'" It would be great if several units could coordinate around the house over a local Wi-Fi or Bluetooth network, while still functioning offline. Ideally, it could be configured from a phone or computer and perhaps have a small wearable unit that could attach to a belt or keys for portable reminders.
Does a product like this already exist, or would it be realistic to commission one? I have limited technical knowledge, so I'm also trying to understand whether this is a manageable project or a much larger custom hardware and software undertaking. Suggestions for existing accessibility tools or suitable technologies would be very welcome.
4 Answers
The individual features are all achievable, but combining them into a dependable, offline, voice-controlled product is more complicated than it first appears. Speech recognition, networking, scheduling, hardware controls, accessibility design, and battery-powered operation each add work. A phone assistant can cover many functions, but it generally conflicts with the requirement that the system not depend on cloud services or use the phone as the primary device.
There are already disability-focused reminder and prompting devices that handle alarms, spoken instructions, or wearable alerts, so it may be worth comparing those before commissioning a completely custom system. A custom build is realistic, but the cost could become substantial because you would be paying for hardware selection, offline voice models, software development, testing, and ongoing maintenance. A local smart-home platform with a few dedicated speakers would probably be the most practical starting point.
Projects based on a small computer and a display or speaker—such as custom smart mirrors—may cover parts of this. There are kits and examples that provide calendars, reminders, text-to-speech, and home automation. You would likely need to combine several projects or add custom software to get the exact wake word, wearable reminders, offline operation, and synchronized units you want.
A local Home Assistant setup is probably the closest fit. With offline speech-recognition and text-to-speech components, it can handle routines, reminders, custom responses, multiple rooms, and control from a phone or computer without relying on cloud processing. The catch is that it takes a fair amount of setup and tinkering, especially if you want reliable voice commands and a polished wearable device.
That seems like the strongest direction if avoiding telemetry and cloud services is the priority. It can be built locally, but it would be more of a DIY accessibility project than an off-the-shelf purchase.