3 ms·

Working on the problem of "blue" light affecting circadian rhythms and sleep. We launched our MVP, Bedtime Bulb [0], in 2018, and it's now the most popular prod

Working on the problem of "blue" light affecting circadian rhythms and sleep. We launched our MVP, Bedtime Bulb [0], in 2018, and it's now the most popular product in the category. We're expanding out of North America to Europe in the next couple weeks. We've had a ton of great feedback from customers, and we are working on several new sleep technologies that we plan to release this year. It's also been interesting to apply the lean methodology to hardware. Iteration cycles are long, but I'd argue that lean is just as important for hardware as it is for software. [0] https://bedtimebulb.com/ https://bedtimebulb.com/

mcraealex16 7y ago
The issue I have with this is there isn't a consensus about what and if any of these products work. The IES (illumination engineering society) has released a couple articles about why consumers should be weary of products that claim to control the rythms and after attending a talk from the IES on the subject I'm not convinced any of it works (at least to the degree marketing says it does) https://www.ies.org/fires/circadian-lighting-an-engineers-perspective/ https://www.ies.org/fires/circadian-lighting-an-engineers-pe...
- yeutterg 7y ago
  I've had a few discussions with the author of that article. Not intending to put words in his mouth, but I believe he is effectively saying 2 things: - The metrics Circadian Stimulus and Melanopic Lux are flawed - We don't know enough to produce products yet I agree with him on point 1, especially about CS. The author, and several others, have demonstrated a huge discontinuity around 3500K. The CS model needs improvement. Regarding melanopic lux, the author says it is more a problem with our measurement and modeling tools. Most manufacturers won't provide a Spectral Power Distribution, but this is starting to change (we do). There is at least one clinical trial in progress that is testing this metric directly, and several experts think it shows a lot of promise. With that said, new research on blue vs. yellow/orange opponency is coming out. Basically, we're trying to figure out contribution of the visual system on top of melanopsin. I'm not sure what to make of it yet, but I think we'll find out much more in the next 2 years. We probably need a better metric than melanopic lux. On point 2, I agree that our knowledge is changing. But it takes a number of years for standards to be developed, much slower than the rate of technology improvement. We are paying attention to all the new research, but we based our design on our best understanding of the research to date. We know that we can really only attract early adopters at this stage who "get" it, and that most people are going to wait and see until things are more standardized. But we are effectively able to do mini-experiments with our customers, which can lead to more insights/avenues of experimentation. Case-in-point: f.lux. Research has shown that it is somewhat effective on its own, but combined with dimming, it could be quite effective. That's actually what we're doing with our product—controlling both the spectrum and intensity. And f.lux is able to run experiments with a portion of their users to advance our understanding. Basically, my opinion is that we have to start somewhere, and we believe we have enough evidence supporting that our solution is "better than nothing." We were able to prove that there is enough of a market for these types of solutions already, even though our knowledge is still evolving. (edited for a paragraph spacing typo and a 2nd time for a grammar issue)
  - germuth 7y ago
    May I also suggest your company to look into the opposite, basically extremely bright lights. I'm picturing one bulb that can go extremely bright in the morning for things like SAD, and then adjust throughout the day. You may think extremely bright SAD lights is already readily available, but there's anecdotal evidence to suggest they are insufficient https://www.lesswrong.com/posts/koRZu53LBZEapwww6/could-someone-please-start-a-bright-home-lighting-company https://www.lesswrong.com/posts/koRZu53LBZEapwww6/could-some...
TACIXAT 7y ago
I have had sleep issues my entire life. I've cut out sugar, caffeine (both for other reasons), and used blue light filtering applications on my devices. None made a significant difference. I think another issue worth looking at is people not getting a daytime light signal. I've purchased and ultra bright light that I saw in this post [1] and it seems to have helped more than anything I have tried before. 1. https://www.benkuhn.net/lux https://www.benkuhn.net/lux
- yeutterg 7y ago
  Yep, reducing blue light at night is an important part of the problem, but it's not a panacea. We realize that everyone has different sleep needs, and much of our focus this year will be helping people understand what inputs have an effect on their sleep.
- Chris2048 7y ago
  Have you tried a cold shower/bath followed by warm bed, or a hot bath followed by a cold fan? Big core temperature changes usually make me feel sleepy.

mcraealex16 7y ago

The issue I have with this is there isn't a consensus about what and if any of these products work. The IES (illumination engineering society) has released a couple articles about why consumers should be weary of products that claim to control the rythms and after attending a talk from the IES on the subject I'm not convinced any of it works (at least to the degree marketing says it does) https://www.ies.org/fires/circadian-lighting-an-engineers-perspective/ https://www.ies.org/fires/circadian-lighting-an-engineers-pe...

yeutterg 7y ago
I've had a few discussions with the author of that article. Not intending to put words in his mouth, but I believe he is effectively saying 2 things: - The metrics Circadian Stimulus and Melanopic Lux are flawed - We don't know enough to produce products yet I agree with him on point 1, especially about CS. The author, and several others, have demonstrated a huge discontinuity around 3500K. The CS model needs improvement. Regarding melanopic lux, the author says it is more a problem with our measurement and modeling tools. Most manufacturers won't provide a Spectral Power Distribution, but this is starting to change (we do). There is at least one clinical trial in progress that is testing this metric directly, and several experts think it shows a lot of promise. With that said, new research on blue vs. yellow/orange opponency is coming out. Basically, we're trying to figure out contribution of the visual system on top of melanopsin. I'm not sure what to make of it yet, but I think we'll find out much more in the next 2 years. We probably need a better metric than melanopic lux. On point 2, I agree that our knowledge is changing. But it takes a number of years for standards to be developed, much slower than the rate of technology improvement. We are paying attention to all the new research, but we based our design on our best understanding of the research to date. We know that we can really only attract early adopters at this stage who "get" it, and that most people are going to wait and see until things are more standardized. But we are effectively able to do mini-experiments with our customers, which can lead to more insights/avenues of experimentation. Case-in-point: f.lux. Research has shown that it is somewhat effective on its own, but combined with dimming, it could be quite effective. That's actually what we're doing with our product—controlling both the spectrum and intensity. And f.lux is able to run experiments with a portion of their users to advance our understanding. Basically, my opinion is that we have to start somewhere, and we believe we have enough evidence supporting that our solution is "better than nothing." We were able to prove that there is enough of a market for these types of solutions already, even though our knowledge is still evolving. (edited for a paragraph spacing typo and a 2nd time for a grammar issue)
- germuth 7y ago
  May I also suggest your company to look into the opposite, basically extremely bright lights. I'm picturing one bulb that can go extremely bright in the morning for things like SAD, and then adjust throughout the day. You may think extremely bright SAD lights is already readily available, but there's anecdotal evidence to suggest they are insufficient https://www.lesswrong.com/posts/koRZu53LBZEapwww6/could-someone-please-start-a-bright-home-lighting-company https://www.lesswrong.com/posts/koRZu53LBZEapwww6/could-some...

TACIXAT 7y ago

I have had sleep issues my entire life. I've cut out sugar, caffeine (both for other reasons), and used blue light filtering applications on my devices. None made a significant difference. I think another issue worth looking at is people not getting a daytime light signal. I've purchased and ultra bright light that I saw in this post [1] and it seems to have helped more than anything I have tried before. 1. https://www.benkuhn.net/lux https://www.benkuhn.net/lux

yeutterg 7y ago
Yep, reducing blue light at night is an important part of the problem, but it's not a panacea. We realize that everyone has different sleep needs, and much of our focus this year will be helping people understand what inputs have an effect on their sleep.
Chris2048 7y ago
Have you tried a cold shower/bath followed by warm bed, or a hot bath followed by a cold fan? Big core temperature changes usually make me feel sleepy.