Zeo shutting down: export your data!

I’ll put a couple of trivial comments on maintenance here to kick off:

  1. If the battery stops charging or the transmitter doesn’t work properly when on the clock or charger, clean the contacts carefully. Particularly if you use disposable electrodes with sticky tape, or put lotion on your forehead.
  2. If the clock display goes blank, you might have set the brightness to zero (particularly if you set different day and night brightness); press the up arrow 8 times. Will probably self-correct by unplugging, but time etc. then needs resetting.

Obviously I can add a note about my software, with download links etc, although that already has its own links

I have two Zeo mobile units and recently purchased a clock only to find that the mobile headband doesn’t appear to work with the clock. I was hoping to use the clock with my mobile headbands to get at the data. I love my Zeo and after trying an exercise bracelet believe that the Zeo is the only accurate sleep monitoring system out there in the consumer market. I am considering doing a home sleep study with professional equipment while wearing my pulse oximeter, Withings Pulse 02 activity bracelet for motion. I also set up an IR security camera to video tape myself during the night. I have sleep apnea and sleep only 4 to 6 hours generally. I have a medical research background and enjoy doing “research” on myself.

I considered doing a cut and paste of all the posts but this seems like a lot of work. Would it be possible to get a Word document or an Excel spreadsheet of the Zeo thread? If so, I would be willing to organize a summary with individual posts referenced as well as any external links.

If I could get the thread in an Excel form, I could sort by user to get individual posts together, eliminate the question posts and hopefully, be left with only the informational posts which could then be reduced in size.

Let me know if this is possible.

[quote=“kurtmyrmel, post:352, topic:561”]
… recently purchased a clock only to find that the mobile headband doesn’t appear to work with the clock.[/quote] Not “doesn’t appear to” but “is known not to”; exactly the sort of information needed in an introduction

I don’t yet know anything about it, but free applications software https://import.io/ claims to do just that.

Other tools that I have used to create a reasonable document from a Web site have to cope with the fact that there is too much stuff that is “rubbish” in a document, though needed in the site. For example, you can have both totally irrelevant stuff (banners etc. etc.) and essential information provided as a graphic; it would be exceedingly difficult to distinguish these automatically. I have made experiments using htttrack, an open-source, multi-platform, 32- and 64-bit program that allows you to download an entire Web site, and working from there. I didn’t find it particularly useful; it just lets you do exactly what you can do on the site, but working locally. If you use it, be sure to get your parameters right; it is possible to set them so as to download the entire Internet, in which case you must be sure to have several yottabytes of storage space available.
http://www.httrack.com/page/2/en/index.html
Docs: http://www.httrack.com/html/fcguide.html

I have also saved single pages (not entire threads) as text from a browser (different browsers produce different detailed text, try several; I think Opera was better than Firefox); this loses graphics information, of course. Or use a word processor (I use Libreoffice Writer, functionally similar to Microsoft Word); either save a page temporarily as a complete HTML page which you edit, or open it by opening the URL in the word processor. This requires a lot of cleanup, and can be problematical (pages can be exceedingly slow, or crash the software). (If importing into Writer, you need to delete all sections, delete comments, and edit links, breaking them all, so graphics are stored directly in the document, not as Internet links.)

For something like a mostly text forum, you can simply copy and paste each posting into your software (word processor document, spreadsheet), skipping some postings (“thanks”), and possibly inserting links to necessary graphics instead of including them in the document.

I haven’t use import.io, but have tried all the other methods, always successfully but usually requiring unrealistic amounts of effort.

HTH

I will consider one of these options but this is the reason that I posted to the administrator hoping that they might be able to provide a csv or excel file.

Gary, it seems to me the major problem with this thread is that, now, the posts are intermingled without regard to the 4 major subjects covered:

  1. General information about the workings of Zeo - theory, methodology, etc.
  2. Retrieving Data from Beside Models
  3. Retrieving Data from Mobile Models
  4. Sensor replacement, rehab, or substitution.

Separating these posts seems, on the surface, to be a daunting task. Maybe, as a start, someone can offer a easy method of separation into these 4 major categories. Perhaps a keyword could be added to each post and then a sort by keyword?

I will say that Woody’s Viewer pretty much takes care of all one needs for subject 2, retrieving data from Bedside Models, at least for me.

1 Like

https://www.keepyoursleep.com/

http://mysplus.com/

It seems that ResMed has purchased the intellectual property from Zeo.

PLEASE, PLEASE I hope I am incorrect.

Does it say somewhere they have? What they are selling on that page is just a motion sensor, like the fitbit or something, but with a display a bit like a zeo, but hardly the same technology.

I was at a meeting 2 nights ago and saw ResMed’s new CPAP machine. I made a comment that it looked like the bedside version of a Zeo (black unit with a pretty blue display which faces the user), just a bit bigger. Someone, not a ResMed representative, mentioned that they just heard from a ResMed employee that ResMed purchased what was left of Zeo.

I asked the Reps who were there and none of them appeared to know what I was talking about. I then heard that the Division of Resmed that supposedly purchased Zeo’s remains is a “consumer” division leading me to believe that this was a different division than the CPAP division.

So this is all hearsay at the moment. What they show on the website, looks like a Dyson product and smells like a bad infomercial. You can make 3 easy payments…
When I clicked on the commercial, it didn’t run but that is possibly because the product isn’t supposed to launch until next week.

When I searched for Resmed and Zeo, I found nothing. I finally found out the unit was called an S+, so I searched for “ResMed AND S+” and that’s when I was able to find the links above.
I did not find the word ZEO anywhere.

I wonder if we should start a new topic for this, so if it isn’t true we can delete the entire topic since this is a Zeo topic.

1 Like

A Zeo employee told me a while back that it was ResMed which bought Zeo’s assets, including the database which I was hoping to get access to. I emailed the guy in charge at ResMed, Colin Lawlor, but never heard back. This is the first I’ve heard of them making any use of the Zeo IP, though. About time.

1 Like

That is the name of the person I was given.
He apparently lives in Ireland but was here in California for the past few days.


Here’s the infomercial. It looks like it radiates the user.

This is incredibly disappointing.

1 Like

I looked at all of this information and it looks like simply another motion detecting device. I have an activity tracker by a company named Withings and it produces a “sleep analysis” that looks very similar based only on the amount that you move during the night.

This is definitely not using the EEG information that Zeo provides.

I use a CPAP and would like to see the CPAP companies add spO2 (oxygen levels), pulse and EEG to their machines. Philips respironics has added a bluetooth module that sends sleep data to my phone and then to a computer website that allows me to track my sleep apnea. I would be nice if I could get these other parameters as well. I wish whoever owns the technology would bring it back though.

Hi everyone,

I am a newcomer to this forum. I note that a few posts ago, a member very helpfully listed the major issues discussed in this thread as follows:

  1. General information about the workings of Zeo - theory, methodology, etc.
  2. Retrieving Data from Beside Models
  3. Retrieving Data from Mobile Models
  4. Sensor replacement, rehab, or substitution.

I am faced with yet another issue. I just bought an allegedly brand new Zeo Mobile device, and it is faulty in that the battery does not (fully) charge (i.e. the green light never stops flashing). When the transmitter is in the dock and charging, I can get just enough flashing blue light time by pressing the button for it to been seen and paired with the mobile phone (and even that took me about 100 attempts), but after that there is no connection.

Am I doing something wrong? If not, can the transmitter be opened and the battery replaced? If I can’t fix this problem I will have to buy another unit. I assume that most units that are for sale will have transmitters that have been lying around unused for a long time, and that may be a cause of the battery going dead and not taking charge.

I don’t know the mobile unit, but it does suggest a bad battery. Hopefully someone who knows the unit will clarify.

Sooner or later somebody will do this and, hopefully, post details; there are details on replacing all sorts of non-replaceable batteries on the Web, but not Zeo yet. You have to work out how to open the transmitter casing, then probably unsolder the battery, work out what it is, and replace it.

There’s a myth that Lithium-ion batteries expire as quickly if not used as if they are; it’s simply not true, battery life is given in number of charge/discharge cycles, though of course they can and do fail prematurely. There’s a great deal of information on this in the Wikipedia Talk page (not the article) on Lithium-ion battery. See the sections:

4 Battery lifetime
7 Battery Life
8 “Battery University”
11 Dubious information originating from discredited source
13 Archiving and unreliable sources
17 batteryuniversity.com

HTH

pol098, thanks for replying.

Considering that nobody has posted about this problem yet (other than here: http://m.youtube.com/watch?v=mKGArcswwOM), and some people must have Zeo units dating back to 2011, it is not a common problem. I might have to take it to an electronics repair shop.

Yes, Lori and gwern are correct, Resmed bought the Zeo assets and is now launching their own sleep tracker, the S+. The official site appears to be at this URL now: http://www.resmed.com/us/en/consumer/s-plus.html.

I wrote to Ben Rubin a while back to see what I could find out about Zeo assets. He didn’t know but connected me to Dave Dickinson, the ex-CEO, who put me in touch with Colin Lawlor at ResMed. Colin confirmed this news, on the condition I wait a few weeks before making it public.

Colin also showed me the sleep sensor they have been working on. Yes, it is a motion sensor. However, the trick they are using is not to directly infer sleep from movement, but to infer respiratory patterns (this is their area of expertise), and then infer sleep from respiration. They are also doing some distance sensing, to reduce noise. The requirement is that the sensor sit on a nightstand, pointing toward the sleeper’s chest. We’ll post some data samples here when we can.

Colin offered to answer Zeo specific questions, or at least try to answer them. Here are the questions I’ve asked so far:

What were the assets of Zeo’s that Resmed acquired?
What did you learn? Anything of value?
Can the “download your Zeo data” be turned back on?
If you aren’t doing anything with the console and headband operational code, could these be released in some way?

He said he or a colleague would answer here directly when he had a chance. Feel free to ask your own questions.

Gary

1 Like

My own question remains the same: can access to the Zeo corpus of sleep data be provided? Ideally as a public dataset in anonymized form, but anything is better than hundreds of thousands of days of naturalistic in-the-wild sleep data disappearing into the void.

  1. Following on from my earlier posts, if anyone by any chance has a spare working transmitter for the Zeo Mobile, I am willing to swap it for a brand new replacement headband with sensors (sealed in original packaging), because it looks like I’m going to end up with two brand new headbands but no transmitter.

  2. Gary Issac Wolf, thank you for your useful contribution. My question to the ResMed would be this: if they are not doing with Zeo’s IP assets, would they be prepared to sell them? Because the niche remains unfilled.

  3. gwern, in what way would aggregate sleep data be useful?

  4. To save me trawling through the entire thread, can somebody kindly answer this question: is there currently a way to export data from Zeo Mobile?

[quote=“Agaricus, post:365, topic:561”]Colin also showed me the sleep sensor they have been working on. Yes, it is a motion sensor. However, the trick they are using is not to directly infer sleep from movement, but to infer respiratory patterns (this is their area of expertise), and then infer sleep from respiration. They are also doing some distance sensing, to reduce noise. The requirement is that the sensor sit on a nightstand, pointing toward the sleeper’s chest. We’ll post some data samples here when we can.
[/quote]

I would be very interesting in seeing how it worked too, especially for those people who move around at night and sleep with a partner whos breathing is louder than theirs!
I don’t really understand how that could work but am obviously happy to be proved wrong.

I suspect there is no chance of them releasing the code, as they will never know whether they will use it, or run the risk of someone else using it.
I suspect they wouldn’t have the ‘download your Zeo’ part as there are big costs in that in that they would have to run some website, but maybe releasing something separately to let people access their data as a one off would be good.
Not much of an issue for me as I can access all the data as I have all the data files from the Zeo, but for those people on mobile or who didn’t keep a copy.

These sorts of databases answer questions as people think of them, so that’s not really the right question to ask: it’s not something you’d know in advance. That said, one can think of many questions if one ponders for a moment: one of my own interests is whether there is a lunar sleep cycle, which there does not seem to be in in the handful of Zeo datasets I’ve analyzed; how many Zeos report highly anomalous data suggestive of the Zeo incorrectly classifying sleep/wake (if you mention Zeo on Hacker News or Wired or Amazon, you’ll be sure to get someone popping up saying ‘the Zeo sucks, I had one and everytime I put on the headband it recorded me as asleep even when I was awake!’); how does sleep vary with seasonal patterns, suggesting the ideal temperature and humidity (I’ve come to acutely regret the lack of a temperature/humidity sensor in the Zeo); what do sleep patterns look like en masse, what is ‘normal’ for sleep?; are there secular trends in sleep indicating sleep deprivation, could we see the influence of smartphones and tablets on making people sleep less and go to bed later? how bad is oversleeping on weekends?