Now this is a story all about how, my project got flipped-turned upside down.
And I’d like to take a minute, Just sit right there, I’ll tell you how I…
changed the focus of my fellowship.
At the start of my fellowship, according to my Bio on the SWCTN data fellow webpage, I set out to “explore new technologies that will enable us to measure this shared experience (non-visual light) “. At the time this was just a step in the journey; it was something I had to do to get on with the stuff I really wanted to do. I would need a way of measuring the light data that was relatively simple and able to measure all the aspects of light that I needed it to. I liked the simplicity of using smart phones. Using the cameras and ambient light sensor showed promise, but only provide basic information. I liked the portability of the smartphone though, so the next obvious step was wearable technology. There are a number of lifestyle products out there, aimed at the health market, that mainly focus on safe sun exposure and Vitamin D. They can be worn like a wristwatch, collecting data continuously throughout the day, giving a cumulative value every 6 minutes, and sending it to your smartphone. I was quite excited to try these out over the summer, and wore one for a whole month. Whilst I got some idea of the my total sun exposure and the amount of Vitamin D I would have synthesised, it was missing the fine detail of all the different aspects of light I was experiencing and and how they changed throughout the day. As with a lot of this kind of technology, the app that it pairs with is just as important as the device, and the sensor can measure far more than the data that the app was showing me. At some point in the future, the manufacturer plans to allow users to access this data, but as these extra measurements impact on the sensor battery life it’s not something that they think the average consumer would want. Apparently my data could be made available for a fee of a $1,000, which has put me off a little. This lead me to attempt producing my own device using a Raspberry pi micro-computer and a series of different sensors. the result was something that worked, but wasn’t exactly as sophisticated or robust as I’d wanted, especially as it needs to be used outside all year round; I had visions of a failed expedition in the Cairngorms, me hunkered down beside a large snowy rock with my handcrafted device in bits! time for a rethink.
Before I had started the project I had bought a couple of teaching grade U.V./light meters for use in the teaching labs at work, the ‘Go Direct Light and Colour Sensor’. I hadn’t really looked at them in much detail outside of the lab class as I knew that they were not quite up to scratch for my purposes. They measured most of the range of light that I was looking at, but the sensors maxed out relatively easily and the analysis software they used was not quite sophisticated enough for my needs. Because of this, I hadn’t really thought about using it for this project, so it was quite a surprise that it played such a big role in shaping just what this project would be. Part of my plan over the summer was to pick a few sites that would give me a good balance of different environments in which to collect some light data. Out of necessity, I used the ‘Go Direct’ meter. I’d picked sites covering open spaces in urban and parkland settings, forests, shopping centres and housing estates. It was a particularly hot and sunny summers day, and I collected data from 11am to 2pm, wandering around each site trying to find the best was to get an average reading across light and shade. When I looked at the data a few days later I found pretty much what I expected. In the open spaces and urban environments the sensors were pretty much maxed out all of the time I was in direct sunlight and all the time I was in shadow they were effectively at zero. In essence the light was too light and the dark was too dark. which is how I’d felt when I was actually there. The only data that wasn’t like this was that collected from forests and greenspaces with some tree cover. Looking at the data, despite the UV index being at levels not normally seen in the UK, the sensors were never overloaded (although they still registered levels of UV high enough to give you severe sunburn), they were also never at zero and there was a great deal of fluctuation of all forms of light due to the movement of the tree cover. The data analysis software didn’t have a sophisticated smoothing algorithm and I was struck by just how much and how quickly this raw data fluctuated during each second. Comparing the two sets of data clearly showed two patterns. All or nothing with very little fluctuation in the bright, open spaces data; or a rapidly and constantly changing set of data in the forests and green spaces. This is the data I want to explore. This is the direction I want my project to take!
What next? Well, I had to bite the bullet and get a specialist spectrometer made to take exactly these kinds of measurements. It’s arrived just in time for Christmas, I’ll keep you posted on what happens next.