Skip to main content

Musings from Philippe

Sailing, Mountains, Music, and Technology

Seasonal Correlation of Heart Rate Through the Night

At Fullpower, we analyzed our accurate multi-year data-set that comprises 250+ million nights of sleep. We found some interesting seasonal patterns. This infographic shows seasonal changes in heart rate. The Fullpower Sleeptracker platform captures continuous heart rate throughout the night completely non-invasively. Each individual fluctuation in the graph is a weekly max and min, the max being in general weekends (bedtime and wake-time discipline are more lax on weekends) and weekdays with a more disciplined schedule and less “distractions”.

This is what we can observe:

  • Seasonal changes occur with lower heart rates in the summer and higher in the winter.  This same pattern was also observed in this independent study in Japan. Our AI-powered analytics discovered this independently and then we found the very interesting Japan paper.
  • There’s a consistent weekly cycle throughout the year with lower heart rates during the week and higher on the weekends (affected by time to bed, diet, and alcohol).
  • We see a big spike for New Year’s eve (time to bed, diet and alcohol).
  • There’s a significant dip after New Year’s, perhaps due to New Year’s resolutions (better diet, decreased alcohol, more disciplined sleep schedule), but eventually, it fizzles.
  • We see another spike after the Super Bowl.
Blog, Technology

Long Before Selfies and Memes, People Wanted to Share Pics

There was a time not so long ago when people snapped photos and didn’t think about sharing them until much later. But these days, you might consider whether to share a pic before you’ve even taken it. Camera phones have made image transmission almost instantaneous, and it’s radically changed the way people take photographs—and perhaps even the way they live their lives.

“You are much more focused on the question of ‘OK, what do I share?'” says Clément Chéroux, curator of the new exhibit Snap+Share at the San Francisco Museum of Modern Art. “It’s not about what I’m going to take a photo of to keep as a souvenir. It’s really about what I’m going to share.”

The impulse to share images isn’t new, though. In the late 19th century, postcards detailing the sender’s location and status crisscrossed postal service routes. As photography became accessible, people subjected friends and family to slideshows; later, they hooked their digital cameras up to PCs and created new albums on Facebook. Still, the speed and scale at which we now express this impulse is unprecedented: 3.2 billion pics every day, each uploaded in a moment, many for a public audience. “It’s not only to one recipient,” Chéroux notes. “It’s to thousands.”

You can trace it all back to a photo of someone’s baby. In 1997, software developer Philippe Kahn became the first person to share a cell phone pic when he soldered cables between his Casio digital camera, Toshiba laptop, and Motorola phone to send his newborn daughter’s face to more than 2,000 people. Within three years, camera phones by Sharp, Samsung, and Sanyo were appearing on store shelves—culminating in the iPhone in 2007 and its game-changing apps the next year. Today the audience is never more than a share button away, and life all too easily devolves into a photographic performance fueled by hearts, likes, and comments.

Snap+Share is an ambitious attempt to grapple with these changes. Among the artists included in the show, Erik Kessels tries to visualize the photo glut in his work 24 HRS in Photos, which is exactly what it sounds like—staggering heaps of pictures representing a single day of all the world’s shares. David Horvitz highlights just how quickly even the most pointless of images spread in 241543903. It features memes—made in response to a call Horvitz put out through his Tumblr—of people sticking their heads into freezers, tagged with a number he made up by combining the serial number on his fridge and UPC numbers on some freezer food.

But it’s the taxidermy cat poking out of a hole in the museum’s ceiling—Eva and Franco Mattes’ Ceiling Cat—that looms the biggest. It’s based on a viral meme of a similar cat accompanied by the warning, “Ceiling Cat is watching you.” Chéroux says it’s a metaphor for surveillance: “If the cat is watching us, the internet is watching us.”

Just something to think about as you share your next pic.

Snap+Share runs March 30 through August 4 at the San Francisco Museum of Modern Art.

Read the original version of this article at Wired.com

https://www.wired.com/story/photography-sharing-gallery/

News, Technology, Technology

Our Sleep Durations Throughout the Week

This week at Fullpower (www.fullpower.com), we’ve been thinking about how much we sleep and don’t sleep each day of the week on average; so we did some distribution analysis. Most of our Sleeptracker (www.sleeptracker.com) sleepers have regulated work schedules which bind them to a fixed weekday schedule. However, there are still several differences. And of course, many of us tend to replenish our “sleep budget” on weekends.

The following image displays the statistically meaningful weekday patterns that we represent using the Sleeptracker AI-powered predictive analytics system.

Technology, Technology

Fullpower AI-pods help the conservation of endangered cheetahs

This endangered cheetah is wearing a Fullpower AI-pod for research purposes. The pod and its associated infrastructure are an AI-powered sensor-fusion edge cloud solution that learns about the Cheetah’s behavior. The AI-pod is small, waterproof, and rugged. We are in the process of learning a lot about cheetah behavior. For example: “When do Cheetahs really sleep?” or “When to Cheetahs really hunt?” This picture was taken at the Cheetah Conservation Fund based out of Namibia.

Technology, Technology

Congratulations 2017 Nobel Prize Winners

Congratulations Michael, Jeffrey and Michael for your unveiling of the molecular mechanisms that control the circadian rhythm.

Yes, our three science heroes have figured out that while we sleep our “protein batteries” get recharged and during the day our “protein batteries” get depleted. It all seems to work in rhythmic patterns timed by the rotation of planet earth. That’s the molecular-biologists’ confirmation of what the data shows, analyzing millions of night of sleep with the Sleeptracker deep learning solution among other things.

Further, the data shows that in eachsleep cycle, each phase of sleep (deep, light and REM), is essential and contributes to regeneration. With every complete sleep cycle, the mind, body and soul get regenerated. On average, for Ms. and Mr. Everyone it takes a total of four sleep cycles to get reasonably recharged and to perform emotionally, intellectually and physically the next day. Using a car as an analogy, think of deep sleep as the engine, light sleep as the body, and REM as the wheels. You need them all, equally, cyclicly, and multiple times during one night, or in separate naps.

From an evolutionary standpoint, genetic research has now established that 25% of us are night owls and perform best at night, 50% are morning larks, and the rest can perform both as owls and larks. What a fantastic opportunity to help Ms. and Mr. Everyone sleep better!

Again congratulations Michael, Jeffrey and Michael for your unveiling of the molecular mechanisms that control the circadian rhythms and for winning the Nobel prize.

Technology

How the Sleeptracker® Cognitive Behavioral Modification (CBM) AI-powered methodology is complementary to Sleep Laboratory Polysomnography Equipment

In a nutshell, the Sleeptracker® Monitor is about small day-to-day non-invasive improvements for Ms. and Mr. Everyone using the principles of CBM (Cognitive Behavioral Modification) that add-up over time to improved sleep in the comfort of one’s home. Polysomnography (PSG) is about wearing a battery of sensors and monitors for diagnosing potentially life-threatening sleep disorders such as severe apnea in one or more sessions in a sleep lab. Both PSG and the Sleeptracker® Monitor are complementary and “compatible.” Big data and AI are what drive the innovation in the Sleeptracker® Monitor.

The Sleeptracker system is focused on Cognitive Behavioral Modification and understanding what simple practical matters impact our daily sleep. The idea is to improve sleep for Ms. and Mr. Everyone with small adaptations suggested by AI working with big data sets. With millions of users and millions of nights of sleep to draw from, the Sleeptracker AI engine helps to give us immediately actionable personal insights as to what impacts a night of sleep. The AI engine observes short, medium, and long-term personal daily sleep patterns and compares the Sleeptracker user to hundreds of thousands of people “just like me” to provide actionable personal insights. Through this process, the Sleeptracker user can experience and quantify for themselves the effects of making measured changes to their sleep routine. The data show key impacts to our sleep quality are directly tied to daily exercise, proximity of meals to bed-time, alcohol consumption and stress. These behaviors which directly affect sleep are handled privately, confidentially and securely for the purpose of understanding and further improving using the principles of CBM. Throughout the night, and completely non-invasively, the Sleeptracker® Monitor continuously monitors breathing rate, heart rate, motion, wake-up and out-of-bed events. In time, monitoring will expand to include air quality, ambient noise and temperature of the sleeping environment to provide greater insight into environmental impacts to one’s sleep. For example, sleeping with a partner, pets on the bed and the impact of children on sleep are important factors which the Sleeptracker AI-engine takes into account. With CBM, the AI-powered coaching agent suggests simple actionable tips that help improve sleep for Ms. and Mr. Everyone, a little at a time.

By contrast, Polysomnography (PSG) is focused in today’s medical world in identifying candidates for CPAP machines. That’s very important and probably the most actionable matter that is derived from PSG study as sleep apnea is a potentially life-threatening condition. One must remember, the Heisenberg Uncertainly Principle (HUP) shows the measuring apparatus of an experiment impacts the outcome of the experiment. Anyone who has participated in a polysomnography study knows how invasive PSG can be to sleep. In fact, often the PSG requires patients be monitored multiple nights to obtain a clear understanding of their tendency to apnea due to the invasive nature of the equipment and restlessness from not sleeping in one’s home environment. Most agree that an important part of sleep hygiene is a consistent schedule and conducive sleep environment. Diagnosing apnea and the prescription of CPAP machines is a multibillion-dollar business and of course is indispensable to those diagnosed. The Sleeptracker® Monitor is meant to be a complimentary product to PSG and not meant to diagnose, but to provide insight into sleep patterns and the measured changes one can make to improve one’s sleep routine.

Correlation of the Sleeptracker® Monitor to Polysomnography is accomplished through a fully operational onsite PSG lab staffed by a licensed sleep specialist. While the focus is on CBM rather than diagnosing apnea, this ongoing calibration ensures the Sleeptracker® Monitor delivers greater than 90% accuracy compared to invasive sleep monitoring. Accuracy is focused on time to fall asleep, sleep efficiency, wake-up events, respiration rates, heart rates and REM sleep.

We’d be happy to have you try the Fullpower sleep lab for a complete free and confidential sleep study if you’d like. Here is an article by Fortune magazine that is two years old and describes our sleep lab and how we use it: http://fortune.com/2015/06/29/sleep-data/

In conclusion, as the data sets grow with Fullpower’s advanced AI methodologies, we learn more and more about our sleep and how the quality of our sleep compares to “people just like me” in a non-invasive sleep environment. Using CBM, we make small modifications that help significantly improve our sleep and health over time. For severe sleep disorders, PSG and a trained MD are key for prescribing a CPAP machine and treating the sleep disorders. For everything else, the Sleeptracker® Monitor offers valuable insight and actionable coaching into one’s sleep routine for a better night’s sleep.

Blog, Technology

Optimizing Your Sleep

What does the data tell us?

The fantastic team of Fullpower Scientists, using the AI-powered Sleeptracker® platform to analyze several millions of nights of sleep from millions of individuals worldwide, dating all the way back to 2011 find that:

– 23% of sleepers are Owls (perform best in the evening)

– 28% Larks (perform best in the morning)

– 49% are pretty much able to perform in the morning, or the evening either way

Morning vs Evening Person: Lark vs Owl

Yes, it’s genetic. If you are not a morning person and can’t fall asleep before late at night or if you love getting up at 6am and feel sleepy by 9pm, it’s not your fault, you may be genetically a night Owl or a morning Lark. No matter where you stand in both of those extremes, it may simply be genetic.

Humans adapt. We’ve evolved from Paleolithic ages by adapting to changes. Yet we are genetically programmed to perform the best in some particular circumstances. The latest research shows a genetic link to being a night Owl or a morning Lark.

A team led by Daniel Katzenberg at Stanford University looked at 410 randomly selected adults to assess their tendencies as morning Larks or night Owls. Among other things they took blood samples and looked at the gene named “Clock.” They concluded that the Larks bio clock was up to one hour ahead on the average to the one of the Owls. https://www.ncbi.nlm.nih.gov/pubmed/9779516

It seems that the world population is made out of one quarter of Night Owls and one quarter of Morning Larks. The remaining half of the population seems to perform equally well at sunrise or sunset. The challenge is that our modern society tends to force everyone to a schedule inherited from the early days of the industrial revolution. This in turn means that there is about 25% of the population that may not perform optimally on a recommended “modern schedule.” You can think of it as lefties being forced to be right handed. Although as humans we can adapt to being ambidextrous, it’s not comfortable. If Kelly Slater were forced to be goofy footed, would he still be the greatest surfer? https://www.surfertoday.com/surfing/7527-the-best-goofy-footers-of-all-time

Owl or Lark? How to Identify your Chronotype

In 1976, the International Journal of Chronobiology published a paper by researchers Jim Horne and Olov Östberg. There is a questionnaire with 19 questions, available on the Internet and published. Here is a link to the original. https://www.cet.org/wp-content/uploads/2014/06/Horne-1976-IJC.pdf

Can you change an Owl into a Lark?

Especially if you are night Owl, society and social life and work schedules may impose stress. Of course, it is always possible to adapt. But it is not advised to change sleeper type. Be a proud Owl or Lark. Don’t force unnatural adaptation. Lefties should be happy lefties and Kelly Slater should not be forced to compete goofy-footed.

News, Technology, Technology

The Creation of the Camera-Phone and Instant-Picture-Mail

Twenty years ago on June 11th 1997, I shared instantly the first camera-phone photo of the birth of my daughter Sophie. Today she is a university student and over 2 trillion photos will be instantly shared this year alone. Every smartphone is a camera-phone. Here is how it all happened in 1997, when the web was only 4 years old and cellular phones were analog with ultra limited wireless bandwidth.

First step 1996/1997: Building the server service infrastructure: For a whole year before June 1997 I had been working on a web/notification system that was capable of uploading a picture and text annotations securely and reliably and sending link-backs through email notifications to a stored list on a server and allowing list members to comment. Remember it was 1996/97, the web was very young and nothing like this existed. The server architecture that I had designed and deployed is in general the blueprint for all social media today: Store once, broadcast notifications and let people link back on demand and comment. That’s how Instagram, Twitter, Facebook, LinkedIn and many others are function. In 1997 this architecture was key to scalability because bandwidth was limited and it was prohibitive, for example, to send the same picture to 500 friends. Today the same architecture is essential because while there is bandwidth, we are working with millions of views and potential viral phenomena. Therefore the same smart “frugal architecture” makes sense. I called this “Instant-Picture-Mail” at the time.

Digital cameras and cell phones in 1997: For the first time in 1997 there was a successful and broadly available digital camera at a reasonable price – the Casio QV. It was 1/4 VGA and made pixelated but nice 320 by 240 pictures. Also, there was a very successful StarTAC Motorola flip phone in the U.S. It was so successful that it had a car speaker kit with a modular plug that allowed for hands-free use of the phone in the car. (That becomes important as part of this invention)

Before June 11th 1997: I had the server-service infrastructure functional as I mentioned. Today we’d call this the “cloud infrastructure”. I also had a Casio digital camera, a Motorola flip phone and a laptop. The camera was designed to upload pictures to the laptop with a special cable. However, there was no practical way to connect the laptop to the StartTac and control batched jobs of for example, uploading a series of pictures to a cloud based system. The laptop didn’t talk to the phone. Yet the phone could send data at 1200 baud over U.S. wireless analog networks. The combination of the Casio Camera, the Toshiba laptop, and the Motorola StartTac were the prototype of a camera-phone that I rapidly integrated on a breadboard thereafter: The Casio Camera because a miniature CMOS imaging sensor, the Toshiba Laptop a small embedded MCU all integrated with cell phone module, both analog for the US and Digital for Japan for example.

Philippe Kahn took the first ever cell phone picture of his then-newborn daughter Sophie in Santa Cruz County.The first “Instant-Picture-Mail” picture shared with 2000 friends and family across the globe: June 11th 1997, Sophie’s birth picture

On June 11, 1997: When we arrived at the Sutter maternity center we had a nice, large private room, with a big desk. We were told that we came “early,” and that we could either go home or stay. We stayed and settled in. I had  the camera, the phone, the laptop and the service-server-system was running at home and functional. It was clear there was a hardware challenge in interfacing the phone to the laptop. I had written a control program on the laptop that uploaded pictures from the camera, used a modem and could upload them to the Service-Server-System with a tag as to which list of friends and family to notify. Then I remembered that I had a StarTAC speakerphone kit in my car. I literally ran down to my car, took out the whole speaker phone kit and started working frantically at creating a software/firmware/hardware interface using the “butt plug” of the Car-Kit to interface the laptop modem to the StartTac 1200 baud modem. My wife Sonia gave me some time because we were 18 hours early. When Sophie took her first breath, all was working and tested. About 2,000 friends and family received the picture around the world. I called it “Instant-Picture-Mail”. We immediately started getting emails and annotations back asking “How did you do this?” and “Can I have one of those?”. That’s when we realized we had something and what we needed now was to integrate the hardware in one elegant component.

Integrating: Immediately it became clear that we needed a CMOS sensor and an MCU integrated into phones. We built these prototypes that were interfaced with the exact software/server/service-infrastructure “Instant-Picture-Mail”.

Integrating, CMOS Imaging Sensor, MCU and Cellular phone modules, late 1997Integrating, CMOS Imaging Sensor, MCU and Cellular phone modules, late 1997

Paradigm shift and Instant-Picture-Mail success: It was clear to me that this was all a game changer and that this was the future of two of America’s most successful companies, Kodak and Polaroid, so I met with their CEOs and core teams. They all had wireless projects, and all could transmit pictures wirelessly and could not envision “Point shoot and share instantly”. No printing, no film, no paper, and no silver halide film. They all had billions of dollars in business, comfortable suites, armies of management consultants, tech consultants and pundits that just didn’t get it. Yes, I was trying to convince them that the future was digital photography “inside the phone” with the Instant-Picture-Mail software/server/service-infrastructure. They hired consultants, market pundits and they all collectively came to the conclusion that phones would be focused on voice (this is before texting) and that cameras would become wireless. Both Kodak and Polaroid went bankrupt and so did Motorola. Nothing could stop this paradigm shift. I couldn’t find anyone in the U.S. with the vision to launch “Instant-Picture-Mail”. I also had no success with my good contacts at big players in Japan, such as NTT Docomo. However, a small Japanese carrier – J-Phone – was enthusiastic and brought in Sharp to design a “Instant-Picture-Mail phone”. In Japanese Sha-Mail means Picture-Mail. It was a great success and it really put J-phone on the map. J-phone is now owned by Softbank. We launched the Camera-Phone with Instant-Picture-Mail in Japan in late 1999.

Return to the U.S. with success: Wired Magazine noticed our success with “Instant-Picture-Mail” in Japan, and in its “Death of Napster Issue” (the one with the funeral black cover), writer Bob Parks wrote “The Big Picture,” cover story which described my creation and some of my IP. The management team at Sprint and in particular two visionary executives: Pierre Barbeau and Danny Bowman, read wired and called me up: “We want to be the first with “Instant-Picture-Mail” in the U.S.” Sprint, Casio and my company LightSurf worked together to launch the first camera-phone in the U.S. in 2002 with the Instant-Picture-Mail software/server/service-infrastructure. This differentiated Sprint, and our partner became the fastest growing carrier for several years. It took two years for ATT and Verizon to launch in the U.S.

What about other claims of inventions: Many companies put photo-sensors in phones or wireless modules in cameras, including Kodak, Polaroid, Motorola. None of them understood that the success of the camera-phone is all about instantly sharing pictures with the cloud-based Instant-Picture-Mail software/server/service-infrastructure. In fact, it’s even amusing to think that none of these projects was interesting enough that anyone has kept shared pictures. You’d think that if you’d created something new and exciting like the camera-phone you’d share a picture or two or at least keep some!

It’s great to reflect 20 years later how the camera-phone is a game-changer for society in so many ways. In the last decade I have assembled the best team of scientists, mathematicians, engineers at Fullpower we are now looking forward to the next paradigm shift. We believe that Sleep will become Digital Sleep focused on measurable, quantified sleep efficiency and performance. We’re pioneering artificial intelligence, machine learning and data science with our Sleeptracker technology to bring a long term roadmap to digital sleep. That’s a game changer for an industry that has been focused on foam and coils for the last 200 years. Just like Kodak and Polaroid were focused on their large profitable brick-and-mortar silver halide empires before they fell into oblivion. Paradigm shifts are both a great opportunity for newcomers and a challenge for incumbents.

Blog, Technology

Sleep Cognitive Behavioral Therapy with Fido

Yes, Fido can teach us humans a lot about sleep. For dogs and humans our scientific approach to analyzing sleep cycles shows us that dogs and humans have similar sleep cycles but that dogs are much more opportunistic in finding opportunities to nap than humans.

We co-evolved with Fido for at least 25,000 years and given the reality of genetics and epigenetics our sleep cycles are similar. Yet our sleep patterns are slightly different. Ballistocardiographs have shown correlation between the study or micro-motions of the human wrist with sleep cycles in humans. That is also true for dogs. When we watch our dogs sleep we observe similar phenomena: Fido twitches her legs as she runs/dreams, snores, growls and smiles through her naps. Fido is really great fun to watch and shows those micro “wrist movements” characteristic of REM sleep and other phases of sleep in humans. Further analysis of brain waves shows similar patterns too.

Of course dog sleeping patterns are a little different than humans. The main differences are in Fido’s sleeping habits. In general Fido is a better sleeper than most humans and a very smart opportunistic sleeper.

Research shows that “Cognitive Behavioral Therapy” is the simplest way for humans to improve sleep, and our best friends can help guide us to better quality sleep. Of course, first we must understand sleep, and there is no better way to understand sleep than to quantify sleep with a solution such as the Sleeptracker.com Monitor. The Sleeptracker monitor gives you a comprehensive daily “Sleep-Score”, analysis and personalized coaching insights to improve sleep quality without changing any of your bedding or habits, completely non-invasively, in the privacy of your home. For us humans it’s essential to quantify first in order to form a good understanding of where we stand and what our goals can be.

And now with a good quantified view of our sleep, watching Fido carefully can complement our quantification with practical cognitive behavioral therapy practical advice.

First question: How much sleep does a dog get (and how much sleep does a dog need?)

Dogs are opportunistic in their sleep patterns. They get all the sleep that they can get when they can get it. They are always ready for any canine future. This is what we can learn from Fido: If there is an opportunity to nap, even for 10 minutes, Fido will. Intuitively dogs feel this. Fido is ready to go for a walk any time, well rested, full of energy. At 5 pm or 3 am, Fido has the same enthusiasm for a walk. On the other hand, we humans are generally not ready to go for a brisk walk at 3 am an get back to bed. That’s because we are all chronically sleep deprived, perhaps partly because we are all missing the opportunities that we may have to “recharge”during the day. Practically, to get some regenerative sleep, we need as little as 10 minutes and 45 minutes is about a complete sleep cycle and wonderful. But we always have “something better to do”. Facebook never sleeps! (But better sleep habits can make us “better at Facebook”!)

Depending on your dog’s size she may need more or less sleep. In general smaller breeds like Border Terriers need less sleep than larger breeds like St. Bernards that may need as much as 18 hours of sleep a day. This article is based on my practical work for the last 5 years with several Border Terriers as well as published research. Border Terriers are athletic and very smart working dogs, bred for their abilities and seem to have a functional and instinctive approach to most of what they do. They are working dogs and they are ready to work hard. And they sleep about 12 hours a day in the aggregate on the average, with limitless energy for hunting rats or taking endless walks/runs on the beach any time of the day or night.

Second Question: How much sleep do human’s need?

Practically as much as sleep as humans can get. We are busy with family, work, hobbies, exercise, Facebook and our sleep budgets are shrinking.

From a CBT perspective we need to learn from Fido:

Be as opportunistic to nap as possible.

From a sleep quality standpoint we need to:

Improve our Sleep Score using advanced non-invasive quantified-self tools such as the Sleeptracker.com Monitor.

By making simple little CBT changes such as eating less before bed and avoiding carbohydrates, drinking that last glass of water 30 minutes earlier, burning that extra stored energy by working out with more intensity (Long leisurely walks are great for the soul but don’t do much for the body) etc…

Yes, one last thing: High Intensity Exercise Can Help Sleep Quality and recovery.

Take a look at Fido, go for a one hour leisurely walk. Great smells great experience, meet other canines and humans. But once Fido gets home, Fido is ready to play ball and run and sprint and do ludic intervals. Then Fido crash into deep sleep. That’s CBT right there. I need not say more!

PS For our Cat loving friends, sleep cycles and patterns are fairly similar. Cats tend to sleep even more than dogs and when cats play and hunt it tends to really be HIIT (High Intensity Interval Training). They are now focused and ready to go at it for as long as it takes and then, deep restful sleep!

Blog, Mountains, Technology

© 2001-2024 Musings from Philippe