A single, binary e-switch is always going to be potentially confusing if it doesn’t simply act as an on/off switch, and more than one mode must be accommodated.
Rather than trying to reinvent the wheel, develop a modern analog slide switch, and put “Off/Low/High” labels next to it, one of the oldest and simplest UI concepts around. No standby/parasitic drain to worry about either. That might come at the cost of a good IP ingress rating, but that’s probably not a concern for grandma. A rotary switch might fare better in that respect, but grandma already has issues opening pill bottles, so it’s a less preferable solution.
For LVP, have it go into blinky mode (at a sedate, non-seizure inducting rate) when the cell runs low, and needs to be recharged. Keep it simple — if the light’s not working normally, that means it’s time to recharge. Grandma isn’t going to be counting blinks and timings like she wants to learn Anduril, and she’s not likely to care how close the battery is to dying. Set a conservative threshold and leave it at that.
It shouldn’t be a difficult goal to accomplish if it is well-defined, and enough effort and resources put into it.
The problem is that any such light is likely to be constrained by cost, pricing, and the size of the market, so to be feasible, such a model will have to piggyback onto the development work for more sophisticated models, and help amortize the costs.
In that vein, a few lights are already in the ballpark. A 3-chip S2+ locked into 100% group 12, or an X2R reduced to a single mode. Give it an OP reflector or TIR if you want to make the beam more benign. Or a simplified version of the SC31, which it should be noted, was discontinued because the business case didn’t support it to Sofirn’s expectations. But perhaps closest, a simpler version of the FC11, which is already a popular option due to its well-rounded appeal.
The holidays provided the opportunity for glimpses into a couple grandparents’ households, and their lights. One was relying on a $10 Home Depot house brand light that I spoke of before — AAA-powered dual emitter, some unknown UI, with run down cells. The other had a Li-powered light at least, an archetypal generic zoomie that now often serves as a shipping container for battery purchases to skirt regulations. Those are the types of lights such a model would be competing against, in price and performance. Or, perhaps the Duracell or whatever lights that Costco sells that are a step higher.
Is there a grandma market for a “real” flashlight, with proper build and performance? No doubt there is, but how large is it, and is it large enough to warrant the effort?
It’s also fair to ask whether grandma wants certain things, like multiple modes, special cells, onboard charging, etc., or are those features are being projected because we like them, so she should as well? Keep in mind the attempt at a simplified Anduril fork, where the amount of feature requests bloated it enough to the point where it wasn’t all that different from the original, full-featured version.