Following a car explosion (not as dramatic as it sounds), our little team decided to hack on in Kent. For the blow-by-blow account, have a look at our Storify http://storify.com/lexij/altfschackkent
We chose to focus on a common field task – quadrats. Fran (@FraMaHa – shortly to be a (not-medical) doctor) walked us through some of the field exercises she’d taught to undergrad Geography and Environmental Science students. We wanted to help students make better use of their time in the field by allowing them to reflect on their results in situ, and benefit from other groups’ data.
We started exploring exactly what students did in the field, and wireframed a possible interface for their inputs. Then we ran through the analysis that would usually happen in the classroom afterwards – for quadrant datasets, these are often biodiversity indicies – and thought about how these could happen in the field in a way that would be useful to students.
Matt (@JasonClint – product manager and would-be coder extraordinaire) started to mock-up a possible front-end, using AppFurnace as a quick prototyping tool, while Luke (@EuriskoStudios – our resident nuclear physicist, CGI artist and programmer) investigated a possible back-end using MongoDB as the data store and Python as the calculation engine. Although you may not split these functions like this in production, dividing these tasks up allowed us to make the best use of the limited coding knowledge and concurrency of our team.
Once we’d got an app that was submitting data, Fran and Alex (@lexij – digital comms bod and general wrangler) went to Darland Banks, a chalk downland SSSI round the corner, and sampled three random quadrats, using the app to submit data back to HQ.
We spent the evening working on the calculations – we’d chosen to use species richness, the Simpson’s index and the Shannon-Wiener index as indicators of biodiversity, as they’re often taught as part of school and university curricula. These are relative measures, so it’s important for students to have more than one quadrat to be able to compare sets of results.
On Sunday, Alex stole kind local designer Allan Willmott (@allwill) for half an hour who helped us convert our scribblings into a logo. We think it looks rather flash
We talked at length about what it is desirable to automate – for instance, we could have automagically pulled in information about soil type into the description field – but part of the learning process for students is to critically observe and evaluate their environment, and make decisions about what information is relevant. We felt it was important to keep the free-form nature of a field notebook.
At this stage, we’ve got a working submission part of the app, working calculations, and Matt and Luke are tidying up how the data is presented back through the app. Fran and Alex are off to the field to re-test.
Future plans (and subject to additional expertise) would be to improve the local storage of data, convert the app to a standalone HTML5 web app using something like jQuery mobile, bring the calculations currently handled by Python into the core web app code, and maybe even add a beautiful, map-presentation of data into the app!