When set out to build Pillow, our goal was to build something beyond the confines of a good sleep tracking application that measures your sleep cycles. We also wanted to provide a high quality resource that will enable smarter decisions to improve sleep quality and gain deeper insight about each one’s personal sleep profile.
Our first step towards that direction was Snooze lab, constantly updated with smart recommendations, experiments and personalised tips. We are already working on its evolution and we are very excited about it. But this post is not about Snooze lab.
Today we are announcing a small but significant addition. It’s an on-line publication powered by Flipboard’s amazing platform. It’s name is ‘Better Sleep’ and it’s a weekly online magazine featuring a specially curated selection of articles related to Sleep Health, the science of sleep and useful tips recommended by sleep experts.
The format is very new and we are still experimenting with it but we really believe you are going to love it. It’s available for free though flipboad, accessible from your favourite mobile device or your desktop browser: here
Take a look, browse it’s content and let us know what you think.
We’ve been kind of silent during the last few weeks, as we were preparing a major update for Pillow. So, after some painstaking effort, the second version of Pillow is finally here and we’re excited to share with you its new features!
In a nutshell, it’s a rather big update including many popular feature requests, adaptations for the latest devices and iOS 8 HealthKit integration and a brand new iPad version. Let’s dive into some more details about each one.
As of version 2, Pillow is a Universal app supporting natively both iPhone/iPod and iPad. Adapting the app for iPad presented quite a few challenges: We had to redesign certain aspects of the interface and mainly the daily statistics screen in order to exploit the bigger screen real estate and perform a variety of adjustments in our other views to make things appear and work smoothly. The incorporation of iCloud gave us an “easy”1 way to sync sleep data between the devices so that users owning both devices can do the recording on iPhone which is easier to accommodate on the bed and browse through their statistics on the iPad, enjoying the larger screen.
Yes, finally Pillow supports HealthKit and interacts with Apple’s Health app. We have already described our intentions regarding HealthKit integration in a previous post and especially how we wanted to include and visualise more aspects of a person’s health and well-being and relate them to sleep efficiency. So, apart from feeding the Health app with the sleep duration data, we incorporated ten more data sources that can be viewed in comparison to sleep quality without needing to leave the app. These sources include: heart rate, systolic blood pressure, caffeine intake, dietary calories, blood alcohol content, weight, steps, cycling distance and Nike Fuel. Every one of these metrics is firmly believed, by current sleep research, to be closely associated with sleep efficiency and be a potential contributing factor or a result of sleep disorder. In that perspective we found it very useful for a person to be able to immediately and easily see and assess this kind of information. We must note that all HealthKit derived data remain on the device and are not stored or shared with any kind of external service or the iCloud, in order to ensure absolute data privacy.
iPhone 6 and 6+
Pillow interface is now also supporting the larger iPhones in native resolution. Since we’re very proud for Pillow’s design we got very frustrated seeing our app in the blown-up compatibility mode and we were truly eager to make things right.
Apart from these major changes, we’ve included a number of smaller yet significant upgrades to the app like the fact that we now offer the opportunity to continue an interrupted sleep session when the interruption is less than 10 minutes.
A short glimpse to what’s next
With the introduction of the iPad version and the HealthKit integration, we believe that Pillow is now one of the most complete sleep tracking apps on iOS. But there are a lot more things in our table for the near future:
Snooze Lab expansion: Snooze Lab is one of our favourite features and it’s clear from the feedback we get that a lot of our users love it. We constantly update it with new facts and everyday tips that can make your sleep routine a bit better and we’ve found out that users find it especially useful. In our plans for the near future is turning it into a fully personalised tool, assessing more available data (such as the ones coming from HealthKit) and providing, in collaboration with sleep experts, of tailored advice regarding sleep health.
Sleep tracking improvement: Proper and accurate sleep tracking is an absolute priority for us and we keep improving and fine-tuning our algorithm in every iteration. User feedback has been extremely valuable in that respect and we will be continuously releasing small (and probably silent) updates in order to better support certain cases such as memory mattresses and CPAP devices.
HealthKit: As we explained above we have already taken the first step in HealthKit integration, but we certainly have a few thins in store. So, one of the first things to come is the separation of “Time in Bed” and “Time Asleep” in the data sent to the Apple’s Health app. It’s a popular request from the users, although it presents certain challenges in the technical level. Still, we feel confident that the respective update will hit the store by the middle of December.
Localisation: We envision to bring support for more languages to Pillow, making it more accessible to people without knowledge of English. Due to the fact the the content of the app (and especially that of the Snooze Lab) is large, we expect to gradually start rolling out new languages at the beginning of 2015, incorporating certain major European and Asian languages.
Celebrating our new update, we are releasing Pillow Premium (offered as an in-app purchase) at a 40% discount and now costs only $2.99 for a very limited time! The download and basic use of the app remains free of course.
As always we would very much like you to try Pillow and send us your feedback, positive or negative. Your opinions and remarks are massively helping us to build a better app.
1 The quotes are in place deliberately since taming iCloud from a developer’s perspective is a true nightmare, but that’s an issue for another post.
One of our great concerns when developing Pillow was how to format our sleep diagrams in order to depict sleep patterns in the best possible way. We had to find a way to combine two different needs: the requirement to show data as accurately as possible and also to provide a diagram that wouldn’t be too complex for the users to understand. Before delving into why we ended up with our sleep analysis diagram, let’s see what happens when a person is sleeping.
How sleep works
According to the universally accepted sleep studies, sleep can be divided in 4 stages and 2 categories, namely Rapid Eye Movement (REM) and Non-REM (NREM). The stages are:
N1 (NREM): It’s the first stage of sleep, where a person is between wakefulness and sleep. Muscular and eye activity is present in this stage and it’s the easiest sleep stage to wake from.
N2 (NREM): In this stage the person asleep becomes harder to wake. The muscular activity slows down significantly but one still reacts to environmental stimuli (noises, light etc). Along with N1, they can be considered as Light Sleep.
N3 (NREM): Also known as Slow Wave Sleep (SWS), it’s what we would casually call Deep Sleep. There is practically no muscular and eye movement and the person asleep is not reacting to most environmental stimuli.
REM: In REM sleep, also known as the dream stage, things are getting more complex. While the person asleep exhibits high brain activity, comparable to being awake, it’s the hardest sleep stage to awake from and at the same time the muscles are completely paralysed. The muscle paralysis is considered to being a safety mechanism in order for the sleeper not to violently react in vivid dreams. Although the precise effects of REM sleep are still under research, it’s considered to be very important for increasing the brain’s ability to learn complex tasks.
During a sleep session, the different stages alternate in successions that are called sleep cycles. The order in a complete sleep cycle is normally N1, N2, N3, N2, REM. A sleep cycle lasts for about 90 to 110 minutes in adults and time in each sleep stage varies, with deep sleep occupying the first hours after getting asleep while REM increases during the sleep cycles before natural awakening. An adult without sleep disorders is expected to spend about 20% of his sleep time in REM.
Sleep Diagrams in Pillow
The most common pattern sleep apps use to depict the different stages during a sleep session is that of a continuous line oscillating between the different sleep stages. To be honest, that was the first thing we tried with Pillow as well, but we found out that there are is a major problem with this approach.
A continuous line has the inherent disadvantage that certain points would fall between sleep stages. But, according to sleep research, sleep stage transitions can occur in an instant. Furthermore, there is no scientific definition for being between two sleep stages: your sleep always falls into exactly one of the aforementioned stages and not somewhere in-between. So, a curved, continuous line would not provide accurate information. It may be suitable for showing the level of motion during the night, but not for depicting the sleep stages. This is in accordance with how Hypnograms look like: what you get is a stepped line and not a curved one, exactly because the transitions between sleep stages happen instantaneously.
Of course, anyone could object that Hypnograms rely on EEG, measuring the exact activity of the brain, something impossible to be done with a smartphone. So, since most of sleep apps measure motion during the night through the device’s sensors, it would be more appropriate to present the level of movement instead of the sleep stages. While this may stand true for many of the sleep apps on the market, it’s not the case in Pillow. In our app, except tracking motion, we use the device’s microphone to detect fluctuations in sound levels during the night, and combine all of our data through a Markovian model to assess the sleep stage a user is currently in. The precision is not (and cannot ever be) equivalent to that of a specialised device but we believe it’s very good given the capabilities of the device.
So, hypnogram seemed like the best fit for showing our sleep analysis. We went a small step further, giving the various stages a different colour, in order to make it even more legible to the average person without sacrificing the quality of the provided information.
Apart from the sleep analysis, Pillow gives each sleep session a quality score as a percentage. Let’s dive into a bit more depth on how this is calculated.
It’s fairly obvious, that the efficiency of our sleep is not always the same. We feel that some nights we sleep very good, waking up rested and full of activity while in other cases even 8 hours of sleep can leave us groggy and unsatisfied. The truth is that sleep efficiency is depending on a large variety of factors except sleep time. Thanks to the capabilities of modern devices we can measure many of these parameters including the sleep cycle duration, the percentage of sleep time in each stage, the disturbances through the night, the environmental noises and the times awakened to name a few. Based on these measurements and using scientifically accepted sleep quality assessment methods, we calculate a sleep score (sleep quality) which is presented as a percentage, since it’s the universally most easy to understand scale.
As we said above, nothing can be perfectly accurate and since self reporting has proven to be of major importance in understanding sleep, starting from version 1.2 we added Pillow a couple of features that will help users correlate the calculated quality with their personal feeling of their sleep. Mood reporting allows anyone to easily record their mood when waking and compare it with the sleep quality calculated by the app. Sleep Notes also help users registering information regarding their sleep that cannot be tracked by the app (for example large caffeine consumption, being sick, being very tired etc). By using advanced tracking in combination with self-reporting we tried to give a tool that will truly give the opportunity to understand sleep much better.
If you choose to give Pillow a try, we'll be more than happy to receive your comments on how our sleep tracking works for you.
One of the new features of iOS 8, and perhaps the one most prominently promoted, is HealthKit. HealthKit, along with Health app, mark Apple’s attempt to bridge the largely disjoint lumps of information about our health that we track every day through our mobile devices or wearables. While there are dozens of great apps for keeping track of our health and fitness status, there is no concise way for them to efficiently communicate and exchange information. Since health and fitness are depending on many factors that are related to and influence one another, having just scattered pieces of information without an easy way to relate them, offers just a fraction of the potential value for to the end-user.
The story so far
When we started building Pillow, our sleep tracking app, we realised that just assessing one’s sleep patterns and quality is simply not enough: we should also provide users with an easy and automated way to correlate sleep with other aspects of their lives and that way help them make more informed decisions on how to improve their sleep quality by adapting their lifestyle. Sleep quality and patterns are deeply linked to many aspects of human life: age, sex, body weight, blood pressure, glucose levels, daily physical activity to name just a few. So, how could we get that kind of information?
A solution would be, of course, to ask the user for it. We could do it, but the drawbacks were many and quite serious: it would cause frustration, require extra effort (imagine switching between three or four apps in order to see and re-input your daily activity to Pillow), effectively ruin the design of the app by adding many extra screens and settings options and could even be considered suspicious in terms of privacy or security.
Since asking the user was out of the equation, our next effort was to find a service that could provide this kind of information to us. The good news is that there are certain services for that purpose. A number of large fitness networks such as RunKeeper as well as Human API already provide ways to interconnect multiple apps, serving as data exchange hubs between them. The bad news, is that to our experience only very few people tend to connect apps there and also that due to the not many developers have apps that support this type of connection.
Despite that, we attempted to make a first step towards this direction by integrating RunKeeper and the M7 processor in the last generation of Apple devices, allowing us to present sleep quality in relation to a user’s daily physical activity. The feedback from our users confirmed that this is indeed something helping people to better relate their exercise to the quality of their sleep and improve their daily routine. Also, we are constantly receiving requests to integrate other apps, wearables and smart devices. But adapting a bunch of different and heterogeneous APIs is a difficult and slow process. The sad fact is that the effort required to properly incorporate even just a couple of the most popular devices or apps is too much for a small team like us.
The road ahead
This is where HealthKit comes into play, serving as the glue between all the different pools of health information residing on our phones. There are two distinct purposes that HealthKit promises to fulfil: the first one is to gather a large variety of users’ health and fitness data to one spot (namely the Health app) while the second, and far more important, is to serve these data to apps, allowing the exchange of related data. Being an Apple service this has certain very strong benefits for both developers and consumers:
1) The adoption of HealthKit promises to be faster and more extensive than any third-party service, since it will built-in and active on iOS 8 from day one.
2) Developers will have a consistent, uniform, native and properly documented way to access a large variety of health data coming from heterogeneous resources, without having to delve into dozens of different APIs to incorporate in their app.
3) Being a native Apple technology developers will quickly incorporate it to their apps in order to gain a competitive advantage. Furthermore, the large amount of iOS users promises that most of the manufacturers in the wearables market will very quickly support it as well.
4) The users will have the ability to easily control data access permissions for all apps in a single point, strengthening the feeling of trust and privacy.
For Pillow we do have ambitious plans for our iOS 8 release. Having a secure and centralised pool of health related information, derived from multiple sources and accessible in a consistent way will give us the opportunity not only to calibrate sleep tracking more appropriately but also to tailor our sleep suggestions and tips to the user’s habits and health condition. And this goes the other way around as well: we will have the ability to export our data to the Health app and make them useful to other apps. We believe that this kind of integration will massively benefit users by increasing the value of the apps they already use and give them a more holistic overview of their health and what they can do to improve it.
We're really excited to announce v1.1 of Pillow: Apart from numerous fixes and improvements in speed and stability, this is the first version of Pillow to include 2 major feature additions: Naps and Mood tracking.
Short periods of sleep, also known as naps, have been shown to improve alertness, memory, motor skills, decision-making, and mood. To help you track your nap's quality and statistics, we added nap detection in Pillow right from the beginning.
Because of the way sleep works, the duration of a nap may have different effects on our body and mind so choosing the right period of time to take a nap can make a serious difference. That's why in this version of Pillow, we've included 3 pre-defined nap modes. "Power nap", "Recovery nap" and "Full cycle" modes. Unlike other similar apps out there, Pillow provides detailed statistics for each nap session and audio recordings.
To access the available nap modes, simply swipe up from the bottom of the app to reveal the sleep modes selection and then swipe horizontally to select the nap mode you prefer.
Tracking your mood after a sleep session is an important aspect of your sleep quality. Pillow now lets you optionally record your mood at wake up time from a scale of one to five.
Your choice is then displayed on the Daily statistics view and you can edit it anytime by tapping on the mood icon displayed for each session. We've also added a new chart in the detailed statistics view, accessible by rotating your device in landscape mode, so that you can view your mood throughout time just as you can do for your sleep quality.