Imagine that your exercise routine is a musical improv band – you jam with it. As you exercise, the band plays improvised music on the fly by reacting to the signals it receives from your exercise data. You react back, varying how you are exercising based on what you hear from the band and how you feel. The flow between you and the band causes you to lose track of time and enjoy exercise in a new way. At the end of the workout, you have a unique piece of music to share with and inspire your friends. You can’t wait to see what the next workout will bring.
For my final project, I’m expanding upon the explorations of Team Xanadu. The general idea is to use real-time exercise data to create generative music with the goal of increasing frequency and enjoyment of exercise.
A basic example is a mobile app that builds upon pedometer data to generate beats, create melodies, and vary volume levels in real-time. As you walk, the pedometer collects data that is used to compose music on the fly. Since the music created is influenced by your real-time exercise data, it will seem somewhat familiar but the generative aspects will ensure that it will never be exactly the same. The idea is that having generative music react to you as you exercise will encourage experimentation and induce a flow state that increases enjoyment of exercise. The main target behavior in the basic example is an increase in the weekly number of steps taken as measured by the pedometer.
At the end of each workout, you have will have created a fun and unique piece of music – an abstract representation of your exercise routine that day. Therefore, another key behavior to target is the sharing of the music that was created. Your friends would probably be more likely to listen to a piece of music than to read your exercise log. The hope is that sharing “exercise improv songs” becomes as common as sharing other music, and that will in turn drive discussion around how they were created and motivate everyone else to increase their activity.
There are countless possibilities beyond the basic system. We could add sensors like heart rate monitors or anything else that collects real-time data related to exercise. We could add more aspects to the generative music – perhaps using recorded sounds from their environment, samples from their playlists, text-to-speech engines, or anything else that has sound. We could also add learning and adaption aspects to the system, so that over time, the system can remember which types of musical elements inspire you to exercise the most and begin to favor the use of those elements when creating the music.
I envision this system being used when people run or walk outside, where the sounds of their environment influences the music created. It could also be used in the gym, where people could jam together and maybe broadcast their music in real-time. The system will have succeeded if people who use it enjoy exercise more and do it more often than they used to.
team xanadu ::
“a fantasy, a musical, a place where dreams come true”
frankie, valentina, and kimi
personal health and team meeting ::
cybernetic model ::
abstracting the cybernetic model ::
exercise and music. “let’s get physical.”
variance and surprise. the lost sense of time. contrex.
participatory design and the meta-designer.
user scenario ::
nytimes blog, “does music make you exercise harder?”
nytimes article “the once and future way to run”
Here are two ideas inspired by the exformation reading.
This bracelet will change temperature based on whether you have enough nutrients. It will range from warm when your body is satisfied with it’s nutrient level and gradually turn freezing cold when you are deficient in nutrients. Basically it serves as a gentle reminder to eat something that is nutrient rich whenever you are lacking in nutrients. As you eat different things and interact with it over time, it reacts to you by changing its temperature rather than informing you of how much protein, carbs, and fat you have consumed. The focus is on nutrition and not on hunger, so if you eat a lot of sugar filled snacks, you may feel full but your bracelet will still be ice cold.
The color of the bracelet also changes depending on what mix of nutrients currently detected in your body, so there is a sense of how your mix changes over time even though you may not know exactly what the colors mean.
These are earphones that change your music selection and volume as it detects tiredness and inability to focus. The idea is that it is a “reverse alarm” that puts you to sleep when it knows you need the rest. So, if you are working on a project after having not slept for a long time while blasting the music to keep awake, and the earphones detect a change in your ability to concentrate, it gradually changes your music selection to a slower paced music and lowering the volume to help lull you into a power nap. After a set interval based on one sleep cycle, it will wake you up by gradually increasing the volume and returning to the music that you had playing before.
I thought this was an interesting take on para-functionality, from a 2007 NY Times article titled Fearing Crime, Japanese Wear the Hiding Place. From wearing what essentially is a costume of a vending machine in order to hide in the urban environment of Japan, to having bag that camouflages as a manhole to hide your valuables, to having a backpack that helps a child transform into a fire extinguisher – this project plays with paranoia and safety in the urban environment. It relates to our fears, stress, and anxiety about the environment we live in – factors that have an impact on our health and wellbeing.
The three feedback system diagrams I created were about lowering blood pressure, increasing vitamin D, and getting a batter out in a baseball game.
It was a difficult exercise because the more I thought about how each system worked, the more complex it seemed to get. It took a some effort to keep the systems simple in order to represent them each as a single first order feedback loop.
The analysis of my first experiment is pretty straightforward since I focused on simplicity. The area of the Fogg Behavior Grid that I focused on was the Blue Dot behavior: doing a familiar thing – drinking one cup of water – one time. I also was really specific about the situation and timing. My goal was to get her to drink one cup of water during her overnight emergency room shift.
The situation had a tremendous negative effect on ability – ER shifts are busy and the extra effort required both physically and mentally to shift away from ER work to even drink a cup of water is apparently very high. I also neglected to focus on her motivation, and my hot trigger was a simple text message “signal” that I sent randomly during her shift.
My girlfriend doesn’t drink a lot of fluids in general and drinks even less water. Since I believe drinking water to be a healthy habit, I am especially concerned when she gets too busy during her 12 hour shifts working at a hospital and ends up not drinking any water at all. So my experiment was simply to get her to drink one cup of water during one of her busy 12 hour shifts.
First, I decided that my hot trigger would be a SMS text message, because it was the least intrusive method – it would not disturb her at work and she could read it when she had a few minutes of free time. I could have done a phone call, but it would have been much more intrusive in her work environment along with a risk that she would not pick up.
I texted her once during her twelve hour shift to remind her to drink some water. When she returned from work, I asked her if she had drank any water. The answer was no. Fail.
My two sketches are here.
The first is inspired by continuous glucose monitors. Continuous blood monitoring with swappable modules to allows one to monitor one thing at a time, whether it’s glucose, cholesterol, vitamin levels, or anything else that you get your blood tested for – and works with a mobile app to remind you and suggest courses of action when your levels are below par and allow you to easily analyze and share your data.
The second uses mobile camera tracking, inspired by Lifelapse, one of the tools mentioned in a QS talk. Instead of using the tool to just log our lives, we could apply computer vision algorithms and have them measure things for us, like the emotions of the people we interact with, how long we interact with them, the variability of our visual stimuli, and even our posture and movement. We can then use this data to trigger actions that will help us improve in those areas of our lives.
I watched a bunch of talks – some taken from the video category of the QS blog, and some from the QS Vimeo page. I enjoyed most of them – in particular I found Seth Robert’s talk on QS + Paleo to be quite entertaining. The two I chose to review were the two that made me think the most.
Robin Barooah had an interesting take on self-experimentation – it was a minimalist approach. He took one weekly objective data point (his weight) and one daily subjective data point (whether he felt lethargic or energetic at 3pm) and formed a habit of recording it over a long period of time. I never would have expected that, without doing any analysis at all, his body somehow figured out what he needed to do to reduce his weight based on just the one subjective boolean check-in during the day. He even threw away the subjective data point, since analysis was not needed to achieve his result.
It’s good to be reminded that data collection is not the primary goal of QS. Bringing your attention and mindfulness to something on a consistent basis – in this case with the daily trigger that prompts you to be mindful of your energy level – can be a way of letting your subconscious do the hard work of making yourself healthier. While it does not increase scientific understanding of health because there is no data to aggregate and analyze, it does seem to be a super simple and effective way to improve individual lives with minimal intervention and hassle. In a world where this worked for everyone, data collection for health would not be important at all – which would be quite intriguing.
Similarly, Nancy Dougherty’s talk was also about triggering the power of mindfulness – but this time with the help of “Mindfulness Pills” that also happen to incentivize data collection by using the placebo effect.
It’s a interesting idea to combine the placebo effect, mindfulness, and data collection into one transparent process. There is a positive feedback loop that is created wherein the user becomes to be aware of their own condition so that they can take the mindfulness pills which make them feel better via the placebo effect. By being mindful of her mental state, Nancy was able to preempt the negative states by taking the “focus” and “energy” pills beforehand and never having to use the “calm” or “happy” pills – which is an intriguing effect. The results seem to work in a similar fashion to the way Robin Barooah’s minimalist experiment did – staying aware ultimately allowed the subconscious to do the work of getting healthier, which in this case resulted in increased mental ability to focus.
In addition, the placebo effect provides an incentive to stay mindful and the “Mindfulness Pills” collect objective data about the user’s emotional state in a way that is minimally intrusive. Similar to how a magic trick redirects your attention – here the user’s attention is focused on getting the benefit from the placebo effect, and the data collection happens behind the scenes. So, like watching a magic trick, I couldn’t help but be delighted. However, unlike a magic trick, there is also a concrete long term effect – the data is then available to be analyzed for personal use or aggregated to derive scientific insights to benefit society as a whole.