How do you feel about this article? Help us to provide better content for you.
Thank you! Your feedback has been received.
There was a problem submitting your feedback, please try again later.
What do you think of this article?
Testing with Friends
I sent one to a friend of mine a few streets away, with no instructions other than 'just plug it all in’, and I asked for a photo - I got this back!
Not only was I pleased by the Flood Alert hardware and software working first time, but after seeing this photo, it really started to make me reflect on how these IoT devices ‘live’ alongside us…
My point being, my friend didn’t test this on a desk, which would feel fine, but somewhat artificial, rather, he’d chosen to place this as a ‘bookend’ - and I really liked the intuitive nature to put this in a place that was perhaps ‘in plain sight’ (ie not in a cupboard, to be missed), but also not in such an obtrusive place that it creates a perpetual angst about flooding.
This is one of those nice moments, where you feel your design ‘leaves your sphere of control’ and takes on a life of its own with the recipients/users of it, and one has to acknowledge and adapt to such feedback, be it pleasing like this, or more problematic, as we’ll come to see later.
Testing With Public Users
Encouraged by this promising start, I posted on my local WhatsApp community group, and Facebook. I got a couple of replies like this…
“You win some, you lose some”, as they say, and although some failed to reply on follow-ups, I did get my first tester, who had experienced flooding in his basement.
This is Alex, who is my first ‘customer’. Looking back, I wonder if this is not a ‘sign of things to come’, in that as much as I’d designed the project to help the vulnerable, probably elderly, and non-tech-savvy, the evident fact is that Alex is not the intended demographic.
Alex worked for the NHS, and he and his wife were both in various aspects of Data Science, so he was not just interested in preventing flooding, but intellectually curious. Although frankly, it was a little disheartening that Alex was not an 85 year old lady, he more than made up for it by being especially engaged with the application of technology for good and data ethics. Perhaps because I’d been working so closely on the tech, I’d assume nobody in the general public would appreciate the lengths we’d gone to to make Flood Alert ethically sound: Firstly, Alexa appreciated that it really was Plug and Play (after emailing me his address and giving permission to film, etc.), and secondly that it came with its own WiFi/SIM - which he instantly recognised what GDPR/Data Privacy/Legal pitfalls it was avoiding, and was highly complimentary about starting off at this more ethically sound foundation. The downside of course is that Alex was the only person to get in touch despite the Flood Alert having a nice little video and being ‘all ready and raring to go’!
Cold Feet on the Tech
I had been in touch with the local group on issues of the Flooding, as most of these people had been flooded, and/or were concerned that the council improves matters to be more preventative (rather than ‘reactive’) towards avoiding flooding occurring in the first place.
Despite initially positive responses from the dozen or so local residents in the early days, I’m not sure if the tech was a bit intimidating, or if the legal paperwork was the thing that gave people ‘cold feet’ about progressing with the project. What would increase the uptake in future projects of a similar nature? Key question to consider…
Evaluation of Flood Alert V1.0 Outreach
Although I can compile a list of technical pros and cons for the Flood Alert, I feel much of the deciding factor about the tech - is how one feels about things. Although every effort was made to keep the legal paperwork in plain English and to be fair about the limitations and liability, I can also respect that for evidently many people, this didn’t feel reassuring enough or perhaps the ‘win’ was not compelling enough to outweigh the perceived downside.
I asked around a few friends of mine in psychology, and also who've run user research and medical pilots to give some feedback and thoughts, in case I was missing a trick...
One suggested that often when things are new, we need champions or sponsors, who we trust to help us trust things as well. It’s often bewildering to weigh the pros and cons ourselves. This is why we have organisations like Which?, through to consumer standards, and CE marks, endorsement, and more besides. I was familiar with Malcolm Gladwell’s work in the Tipping Point, and he described these folks as ‘Mavens’, in that they are ‘trusted sources’ in your social groups.
Such ‘sponsors’ or ‘mavens’ are of course hard to initiate with a prototype and to do a small-scale pilot. And perhaps a failing was to not offer money to take part in a trial…? Although most commercial trials are paid, because the company is the ‘winner’ in the deal, this project is peculiar as it is not intended to be a commercial product, (open source), and the ‘win’ is helping the user to gain more awareness of their predicament. With that said, perhaps people felt that the ‘win’ for RS, and indeed myself, was nonetheless gaining favourable PR.
Another friend and mentor empathetically pointed out that ‘this is one of the rare occasions where celebrities are genuinely useful - so long as they are legit!’ - their point being that they ease our decision-making process, as they essentially ‘vouch’ for the trustworthiness and validity of a product or service. I didn’t quite see how I’d be able to get Brian Cox or Hannah Fry, or the like to give Flood Alert the ‘thumbs up’, but at the same time, I totally saw their point!
A New Chapter: Going Back to Go Forwards…
For us, the goal for Flood Alert is still to offer help to those who are less tech-savvy. Nevertheless, it looks like we might have some ‘credibility-building’ to do before it can be rolled out widely!
Recognising that the folks who have taken the most interest in the Flood Alert project - are those who work in Data Science (5 people); AI and Machine Learning (8 people); and general tech and design (25+ people). I had initially dismissed these connections when I came across them in professional circles, as they were not in my ‘target’ of ‘vulnerable’ and ‘non-tech-savvy’. However, upon reflection, I think we might have to do a Pilot with the ‘tech savvy’ - first - and then once the project has garnered more public attention, been seen to be overwhelmingly trustworthy, and perhaps if not Brian Cox or Hannah Fry, then at least a few credible folks who’d endorse it, perhaps that would be compelling to the group of people I’d like to help most.
The Benefit of Hindsight
Initially, the Machine Learning part of the project, as crudely illustrated above, was intended to be a ‘Part 2’ following on from the initial launch, with no ‘Machine Learning’. This was because this seems really intimidating: Asking a 85-year-old person to participate in a ‘Flood Alexa’ project felt like they might possibly roll with the pilot, as the ‘Alexa’ metaphor felt familiar. However, ‘AI’ and ‘Machine Learning’ is intimidating to most folks, even those much younger than me, who are not well-versed in the tech industries or other technical disciplines. In short, I thought adding all the AI/ML parts right away was too much too soon!
However, now that it looks like we’re effectively ‘re-launching’ the project, but this time targeting the ‘Pete Milnes’ of this world, who have been affected by flooding, but who are confident with tech, this suddenly is less of an issue.
Furthermore, when I first drew this concept, I was aware that the Data Privacy issue was going to be tricky, as I had assumed I’d be using the User’s WiFi. However, as this ‘reboot’ of this project was now going to have its own dedicated WiFi and SIM Card data plan, this radically changed this aspect now, for the better!
Focussing the User and the Flood Risk
Now that I had decided to ‘zoom in’ and focus on the right ‘type’ (techy, and affected/worried about flooding), I also realised I needed to be more specific also about what the Flood Alert can realistically add benefit to.
Being totally honest, those in Wood Street who suffered 1m+ flooding, my Flood Alert will really only tell them to perhaps ‘brace for impact’, and no amount of sandbagging will really save them, unless significant drainage and infrastructure improvements are made by the Council and Water Authority.
However, those who suffered from the ‘micro flash floods’ - those situations where one needs to sandbag to protect against a sudden rise in surface water, which usually goes down again in an hour or two, but if neglected causes 6 inches of water to enter your home, and the damage to flooding, furniture, electricity, belongings, etc. is still very problematic, is what the Flood Alert is likely to be most useful for. Perhaps it should be called a ‘Street-Level Flood Alert’ or ‘Urban Flood Alert’...? This also linked-back to discussions with David from Thames Water pointed out (and indeed, seconded by residents in some parts of Walthamstow), as well as anecdotes from Flood Forum.
Perhaps by being less ambitious, I’d have more chance of actually defining a credible intervention which was not unrealistic to be helpful, but to which many people still would appreciate not only the Alert, but the advice to cope better.
The Serendipity of a New Role at the Royal College of Art
At this time, I had also joined the Royal College of Art Ecological Citizens team as their Technologist in Residence, with the role being to explore how tech can help with citizen science. Sometimes things are serendipitous!
One thing I have learned from EC’s noteworthy My Nature Watch project - which involved getting people to place mini webcams near bird feeders, to observe ecological insights about birds, was that the project didn’t just materialise ‘out of thin air’. Rather the project was advertised to birdwatching enthusiasts - an already very dedicated and enthusiastic group of people (often with a bit of disposable cash for said hobby)... and from that there was a subset of people who’d be willing to ‘tinker’ with a bit of tech, in this case, a Raspberry Pi computer and a mini camera, and ‘join the fun’. Indeed, I not only had a go at this myself, building my own variation, (link) - but in the process met an engineer on LinkedIn who worked for Intel’s Machine Learning (GETI) team, and whose hobby was Birdwatching!...
Sometimes you go from feeling ‘You can take a horse to water, but you can’t make them drink’, to ‘birds of a feather flock together’! So this was a teaching moment of how ‘finding the right crowd’ be it birding or flood prevention, was key - and then finding the subsets of people who wanted to explore how tech can help. For example, to visualise this…
What I realised in hindsight was that what I was basically describing was how easy it was to ‘nudge’ a group of people to try something unknown, but which they are already motivated/emotive above. Also, the relative proportion of folks within the subsets - anecdotal I think the slight ‘nerdy’ nature of birding tends to attract folks who are into tech also. Conversely, those suffering from flooding are more like the ‘general populous’, so the proportion of techies is smaller, relatively speaking. Yes, there are more of them, (general population), but in the busy world of social media, you need things to spread word-of-mouth for these sorts of projects, as it is the ‘Connectors, Mavens and Salespersons’ that Malcolm Gladwell describes in The Tipping Point - who we needed to engage in this re-boot.
Granted Birding is fun, and Flooding is not, but my inclination was that I needed to connect with people who were somewhat already proactive in the matter of flood prevention, comfortable with a bit of tech (or at least curious), and who had time to invest to get the most of out of the pilot. The motivator is of course getting a better prediction of when you might flood.
Given Heather’s discussion about the stress of the uncertainty, (Flood Forum interview), I hoped that the Flood Alert (with AI/ML) might also grant those at risk of flooding a greater sense of agency.
Sometimes the answer is right under your nose…
The funny thing to realise is that the whole reason Pete Milne even joined in on this project, even before it was formally commissioned by RS DesignSpark - was that he cared about Flooding and Tech.
Pete has worked in tech for years and is always curious about new possibilities. Pete has also sadly been affected by flooding in his home, and since has moved home, and although he’s thankfully not in the same predicament now, he nonetheless has evident sympathy toward those who are suffering from (or anxious about) flooding and has many practical and empathetic insights into the situation as a result.
For me, Cumbria also remains emotive, even if I am not directly affected, I know family and friends who have been, or who are still battling with the stresses of it. Although this project was trying to be efficient and local to Waltham Forest, it may be that it needs to cast its net further afield to recruit around the country, then engage a specific group to help establish the core technology, before reverting back to helping as many people as possible, from all walks of life.
And so the next chapter - with your help also begins!
Please do get in touch if you’re fairly technologically minded and have been affected or are worried about flooding, or indeed. Alternatively, if you’re in a position to be the ‘conduit’ between us and someone less capable or confident, perhaps friends or family who need help, we would also love to hear from you!
Footnote: From Walthamstow to UK Wide Availability
In case you were wondering why I was so fixated on Walthamstow, and not the UK, this is also something which is ‘fixed’ by ‘rebooting’ this project - with the benefit of considerable hindsight!
When I began this project, the assumption was that because I’d be helping vulnerable people, who were not techy, who were at risk of flooding - I would have to be their ‘tech support’ and likely have to pop around to help for some very basic tech issues.
If I had a pilot with say 10 vulnerable people, imagine having a call-out say 2-3 times, and that’s potentially 30 train tickets around the UK… which was just not realistic for my time, and the project budget. So hence the Waltham Forest borough, where I lived, was convenient to run a pilot, and in certain areas did indeed have flooding troubles which were not just river-related (Fluvial), but surface water-related (Pluvial).
Of course, now that the end-point was to create a truly ‘Plug and Play’ setup, combined with the re-focussed pilot group now being ‘techy’, this means if I send a kit (above left) via Post to a ‘Pete Milne type’ of person in for example - Scotland, this is okay, even though I am unlikely to travel to support the beta testing, they would be likely to be able to handle the technical nature of it, and perhaps then some. With this re-focus of users, I might be able to do some quite complex AI/ML stuff, which will require firmware updates and such, and which would frustrate non-techy folks, but we can figure out all the issues with this group.
Blog Series Contents:
Prologue - The Case for 'Hyper-Localisation' of Civic Data
Research & Development:
Part 1: Filling the Local Data Gap
Part 2: Civic Services & User Experience Research
Part 3: Ideation of Flood Alert Concept
Part 4: Prototyping Back-Story
Part 5: Citizen Science Learnings
Open Source Build Guide:
Part 6: Build Guide for 3D Printed Assembly
Part 7: DIY Decals for 3D Prints
Part 8: Code & Data Guide
Future Ambitions:
Part 9: Project Reboot with Machine Learning
Comments