We are all familiar with the concept of reinventing the flat tire in digital development. That’s when we repackage a known failure as new innovation, usually for a new donor or government Ministry.
What about the opposite though? Reinventing the digital wheel. When we repackage a known success as novel innovation to a new buyer.
Reinventing the Digital Wheel
Benita Rowe’s comment on another ICTworks post inspired me to think about when we resell ideas already proven elsewhere as a new innovation in a new context. You could say that it’s a valid innovation. One definition of innovation is the introduction of something new. It doesn’t have to be new in an absolute sense, just new to that situation.
The use of drones in agriculture can be classified as innovative when they are deployed to monitor crop growth in geographically vulnerable contexts. You can also call it replication. You are truthfully replicating that innovation when you deploy it in different locations.
However, I am always surprised at how often I see well-tested solutions touted as somehow a groundbreaking innovation. Not innovation in a new context, or replication. Just pure innovation. This is particularly common with donor-funded interventions.
Internet in a Box for DFID/FCDO
Recently, the Technology Frontiers Livestreaming program released a flashy brochure about an Internet in a box pilot they did with DFID/FCDO. “Internet in a box” or “School in a box” solutions are usually an educational content server, a wifi access point or wifi router, and power supply in a sturdy box.
Internet in a box – or any digital content distribution solutions for education – should not be described as an “emerging technology”, nor should it be a question if they can provide education access for children in Internally Displaced Persons camps. The technology is already proven and the solution widely deployed.
- In 2000, eGranary started deploying content servers across multiple countries, and into purposely offline locations, like prisons.
- In 2015, Inveneo launched Solar Libraries that used solar panels to power the content servers in off-grid, offline locations.
- In 2017, Ustad Mobile’s app turned any Android phone into a wireless offline content distribution system that was deployed in every public tertiary institution in Afghanistan.
What we now know and do not need to keep rediscovering: access to digital resources on [insert the device of your choosing here] can increase learning gains, providing additional factors such as ICT competencies, nutrition, the regularity of access and time on task are controlled for.
However, in many contexts, an offline-first approach to educational technology – one that makes use of the mobile devices that people already own – and equitable access to distance learning are inextricably linked.
Internet dependence, extra hardware or infrastructure, corporate social responsibility partnerships, or the regular physical presence of users in a public setting are often neither viable nor financially sustainable options, particularly for children in IDP camps during the COVID-19 pandemic.
3D Printing Hands for CISCO
Private sector companies like CISCO are not immune to shiny flashy technology pilots. They are just as susceptible to captivating technology narratives as large bilateral donors.
Recently, Mercy Corps posted about their 3-D printing pilot in a refugee camp that was funded by CISCO. The pilot sought to develop assistive equipment for children with physical challenges.
After many trials getting the equipment into the camp, training operators, and running the printers on a regular basis, Mercy Corps was able to 3D print prosthetic devices. Plastic hands that were first designed in 2012.
- In 2014, we learned when 3D printing was useful for humanitarian programs through an IREX Deep Dive with the founder of a 3D printed hands organization.
- In 2015, you could get certified in 3D printing for international development organizations with a TechChange online course.
- In 2017, DAI showed when 3D printing increased resilience of communities facing humanitarian crisis, including the limitations of small-batch fabrications.
Most frustrating is that Mercy Corps relearned all the issues with bespoke additive manufacturing that we’ve known for years now. Things like computer aided design complexity, machine capacity limiting the size of fabricated objects, and the need for different filament printing materials.
Blockchain Systems for USAID
With the price of Bitcoin reaching new highs, it will only be a matter of time before a USAID implementing partner publishes their own effusive marketing about a blockchain pilot as innovation. Blockchain for development is not an innovation. In fact, few things called blockchain are really blockchain public distributed ledger technology.
We do know which blockchain implementations work for humanitarian organizations.
- Land tenure is a great blockchain use case – after completing the messy process of identifying who owns what creates an initial starting point for the distributed ledger.
- Digital identity is another use case where a public immutable record can help people and organizations know who is, or isn’t the person they say they are.
- Financial services is an obvious use case – see Bitcoin – and this can be for any payment, be they payments to an organization’s constituents or remittances between private parties.
I only worry that the Bitcoin hype will mean we see failures repeated. We already know that blockchain identities for refugees is a really bad idea. We also know that cryptocurrency donations don’t work. As I wrote about before, not even donations to UNICEF are successful. Worse, when blockchain efforts were systematically analyzed, researchers found 43 pilots and no impact.
It’s imperative to design technology-supported programs that build on the knowledge we already have. Indulging in our guilty pleasure of reinventing the digital wheel for new donors disadvantages those most vulnerable. In an effort to do good, we must be vigilant to do no harm.
Internet in a Box for DFID/FCDO – I want to add to the list TECHAiDE’s ASANKA (www.myASANKA.com) in 2017.
Thanks Nadia, such a spot-on and well illustrated blog post! Definitely recognize the guilty pleasure – examples are abundant.
My alternative definition for innovation continues to be: “A word that you start randomly adding to your proposal when you have ran out of ideas and want to impress the potential donor”.
Hi Nadia,
I really enjoyed reading this post, thank you Nadia. I think especially because I wanted to engage with it almost immediately and challenge some of the premise. I would be happy for a longer discussion, or please feel free to provide further dialogue here, as this is something I’m really interested in, and I think you bring up absolutely valid points. However, I’d love to take an “enterprise” look at this. I was thinking that one ‘innovative’ solution, done once, doesn’t mean that a second solution isn’t necessarily an innovation, even if it uses similar tech, similar premise, similar location, etc, or even different tech/premise/location.
For example, Microsoft created a computer. So did Apple. So did a lot of other companies. Same as smartphones, etc., Samsung, Nokia, Huawei, Apple, htc. And I would say that to become a successful enterprise, you can’t just produce exactly the same product as a competitor has created. You’ve got to innovate a little, put your own spin on it. And see how the market responds – does it do better, worse, what tweaks can you make based on what you learn?
And so, actually, maybe what we’re not great at doing in the sector is recognizing “what” the innovation is…for example, internet in a box. Okay, maybe the tech isn’t groundbreaking – but what about Internet in a Box is substantially different from previous iterations that make it 1) more adopted, 2) more impactful, 3) more useful, etc, etc. Is it the marketing to local schools? Is the communication better? Is the customer service to remote schools that’s better?
I’m not sure if I’m making perfect coherent sense – but I’d love to see WAY more field teams latch onto technology, understand what worked and what didn’t, and then innovate their own spin to make it better.
Or perhaps the challenge is getting donors to push better for explaining “what” the innovation is. For some reason, we always seem to want to think that the tech itself needs to be the innovation, whereas I think the innovations that are the most interesting are “how you use” the tech – and there are sooooo many ways to twist and turn this.
Potentially I’m thinking more from a social enterprise perspective, but also, in my work with remote humanitarian field teams, I see the exact same questions being asked as I asked ten years ago about the same technologies. Our sector has such massive turnover, at such a high rate, that building the institutional memory is difficult, and so we do need to keep not just “recreating” the digital wheel, but recreating it better and better. I’d love to see more of that happen across programmes.
Hope some of this made sense? Excellent article, thanks for inspiring!
Janna
Great post! It’s it’s worth mentioning that this industry has fairly perverse incentives to brand everything an innovation as opposed to investing in or promoting well proven tools and approaches. We have to pitch everything as a novel technology silver bullet, but we all know that digital development is fairly incremental. It needs to appear that the vast majority of our budgets go to R&D, where as in the private sector, where most innovations are derived, only a small percentage of a corporate budget actually goes to R&D. I would argue that succeeding with an existing technology/global digital good in our space IS innovation and novel. The technology might be proven but the adoption, sustainability, training, business model, etc to a new context is often where the most innovation is needed. It’s not too sexy though to say we developed a new adoption model for a old technology. My biggest concern is that this need to reinvent the wheel is actually an artificially generated market with in the development sector. In my experience there is fairly little country demand for this. Many countries, like most of us, want proven solutions they can have ownership over, trust in, and can institutionalize into routine processes. Well timed and placed disruptive tech solutions can break paradigms and bounce the sector forward, but when all you perceivably have are paradigm breakers its ultimately just seems like a lot of noise.
Scott, I think you have a great point about who wants these innovations. The trend seems to be donor driven much more than country government driven, who, like you say, want proven solutions they can manage directly.
“The technology might be proven but the adoption, sustainability, training, business model, etc to a new context is often where the most innovation is needed. It’s not too sexy though to say we developed a new adoption model for a old technology.” – I think that’s one of the most challenging realities when fighting for funding – where real innovation is needed doesn’t sound quite as exciting. We need to get reeeeaaally excited about operationalising (is that a word) the tech into super-contextual, relevant ways. Huge agreement with your post here Scott.
While i concur with what you are saying, i also think that you are missing the fact that off the shelf tech solutions are NOT designed for humanitarian sector. You know, why don’t we patent “humanitarian action-by-design”?
Precisely for the reasons you list: “adoption, sustainability, training, business model, etc” but also, and make no mistake it is a big endeavor, to be capable of standing straight in your boots an claim “we did everything we could to minimize harm” and not look like those guys in the Social Dilemma apologizing clumsily with “when we deployed tech X Y Z, we did not imagine these bad things could happen”.
This does mean R&D investment and, for some, re-inventing the wheel. Did Signal re-invent the WhatsApp wheel?
☺️
Dear Nadia,
I share your exact perspective and I have been working on something that I feel can be a step towards a solution. Do you have an email or linkedin to continue discussion?
thank you
Nir
This is a thoughtful article about “innovation” and assumption of innovation from existing technology. I hesitate to fully embrace the full narrative though because to me, innovation includes new applications of existing technology. The first example discusses 3D printing for prosthetics in a rural village.. The tech is well established, but the application is new. It is our responsibility to not just design and develop innovation, but also to implement broadly.
There is definitely some validity to the topline point that the author is trying to make, although I did not find the way it laid out here to be compelling or entirely accurate. The post is predicated on the point that it is bad to make use of something that is known to be a success while claiming it is something new and innovative.
Yet, if you look at the examples cited, such as the FCDO or Mercy Corps documents, they don’t actually claim that these ideas are cutting edge or innovative. Also, is there really anything wrong with taking pride in doing something or sharing what you learned just because someone else has already done it first? If you are going to call people out for doing something, then make sure they are actually doing it and that doing it leads to negative outcomes–or if not, find better examples.
The post defines the key term in its title as “repackaging a known success as novel innovation to a new buyer.” Yet it ends by saying “It’s imperative to design technology-supported programs that build on the knowledge we already have.” So which is it? Is it bad because practitioners are overhyping things that work, or because they are not paying attention at all to the lessons learned in the past and muddling their way into learning something that many before them have already learned? Certainly both of these points can–and sometimes are true–but confounding the two in this post takes away from the overall argument.
The author also makes a number of claims related to things that are definitively known to work or not work. For example, she states that “We already know that blockchain identities for refugees is a really bad idea.” The reality is that the source cited here is just an opinion piece by Wayan, not authoritative research or practitioner consensus. As much as I like Wayan, just because he says this is a bad idea it does not mean that “we already know” it. (I’m not arguing it IS a good idea, just making the point that one person’s opinion is not the same as wide consensus or definitive proof).
Despite the many shortcomings I find with how this post was laid out and the evidence and analysis used to support it, at least it has led to a robust debate here in the comments, which is a value in itself.
Note: I write all of the above in my personal capacity, and not in my official one. They are my opinions alone.
Hi Josh, not to reopen what I hope is a closed debate, but exactly when do you think it is a good idea to share an already exploited and persecuted population’s identity in an immutable public database? My opinion on forced blockchain identities for Rohingya refugees is not based in authoritative research (thank god!) but it sure is practitioner consensus. Not everything needs an RCT to be “known”.
@Wayan – we’re talking about two different things here. I’m not arguing that forcing Rohingya refugees to have their identities put on a public blockchain is the right thing to do. But the author didn’t state that. She said “We already know that blockchain identities for refugees is a really bad idea.” That’s a blanket statement that applies to all types of blockchain-based identities, including those that are on a permissioned (i.e. no public) ledger and in all situations. There is certainly no consensus on that being always bad in all cases. I also agree with you that not everything needs an RCT, that’s why I said her statement is not backed up by either authoritative research or practitioner consensus. I stand by that.
@Josh – when tryng doing no (or minimum) harm is the driving principle, if there is no strong supporting element in favor of deploying a given tech then don’t do it.
In other words: if you cannot prove that DLT is THE solution that solves a identified problem and knowing the risk of that tech (whether technical or due to the current hype), i suppose you should not do it.
You seem to start from the opposite asking to prove it is risky.
Or i am misreading?
@Vincent – you’re misreading me here. I agree with you. I think we do have an obligation to make sure that any technology is appropriate to the context and needs AND that we seek to do no harm. My point here is only that it is not correct to say that everyone already agrees that digital identities for refugees that use DLT is a bad choice in all circumstances and contexts. Maybe there are really no instances where this is the best solution, but no one can claim right now that ‘we already know’ that to be a fact.
Hi Josh,
The first example Nadia cites is an education project that experimented on refugees during a global pandemic – without considering what has already been done in the space and building on that knowledge – while claiming to be cutting edge and innovative. I’m not sure why she’s being told that she’s calling people out without evidence when she has included a link to a report for that program in the post. Drawing on that example specifically, I would disagree with you and argue that there really is something wrong with people taking pride in sharing something that they have learnt – something which is already long-established and did not need to be rediscovered – at the expense of people in an extremely vulnerable position. What Nadia is calling out here are very real power disparities in the humanitarian sector that do harm. The use of “repackaging a known success as novel innovation to a new buyer” and “It’s imperative to design technology-supported programs that build on the knowledge we already have” do not take away from her overall argument as both are sufficiently addressed in the post and the former is supported with a linked report. Finally, it’s a blog post format, not an academic article and should be critiqued with that in mind.
@Nadia – I very much enjoyed your post – well done and thank you!
@Benita – did you read the document that she linked to? Nowhere in that document do they claim that what they are doing is cutting edge. That was the author’s own declaration. There are two references I see to innovative or innovation in there. One is a quote from someone on their team who is from Myanmar. Considering that widescale access to mobile phones in Myanmar is less than a decade old, you might forgive him for feeling that this is an innovation. The second reference is in the context of where an innovation in the context of Myanmar would influence the government’s approach. If this has never been done in Myanmar before, then it is accurate to call it an innovation within the Myanmar context.
If they are taking pride at the expense of a vulnerable population, then the author should have made that case. I don’t see anything in the FCDO document that would lead me to believe that they were somehow exploiting the group of students who were in the pilot. If you see otherwise, please point it out.
Last, I’m not critiquing this as an academic article. I’m critiquing it as a blog post. Just because something is a blog, it doesn’t mean that we should play loose and fast with the facts. Strong statements should be backed up by strong evidence. If you can’t back it up, then tone it back. As I said, I think that there is some validity to the author’s points. I just think it was presented in a way that was full of holes, which ultimately does it a disservice.
Josh…debating the frequency of words in a document is beside the point. For what it’s worth, yes, I have read the document Nadia linked to and my hair is still standing on end as a result, months later.
On the rest of it – my opinion remains the same for the reasons I outlined above. I think it’s a great post that addresses some real issues and backs these up with valid examples.
Let’s agree to disagree on this one – cheers.
@Benita, I’d love to know what about that FCDO document has your hair standing on end. Perhaps I missed something that you could elucidate.
(can reply more above so starting here)
@Josh: i think you are being a bit pedantic here as, formally, i indeed do not see how “[…]everyone already agrees that digital identities for refugees that use DLT is a bad choice in all circumstances and contexts” can ever be reached.
Isn’t there a generally accepted (see i am improving ;)) that when you are “beyond reasonable doubt” you can call it a day?
In this case, and sorry to disagree with you, we are today “beyond reasonable doubt” that DLT is not a good idea for DID for most vulnerable populations (e.g. and only from a technical stand point: connectivity requirements, cost to run/maintain, subject rights for rectification, complexity to effectively have your copy of the ledger, lack of standards of the DLT architecture/consensus, DLT choice lock-in, security of the endpoint app/wallet; then add legal, cultural acceptation, surveillance risk, etc)
Again, i would be please to review my judgement could you come up with one context where THAT particular DLT based design does solve properly a real problem.
@Vincent, I don’t think we need to get to uniform consensus. Most of us, hopefully, operate using “beyond a reasonable doubt” for our decisions. Waiting for clear evidence of harm would be irresponsible if we know that there is a likelihood that harm could be caused. I think we would agree there.
Going back to my original point, the author points to an opinion piece by a single person about a single use case as proof that DLT for DID for refugees can never and will never work (or at least to strongly imply that). Is it pedantic to expect that people do not make blanket statements without sufficiently backing them up? If so, then I am guilty. 🙂
I don’t have a horse in the race here when it comes to DLT. I’m not advocating FOR or AGAINST it in this context, as I don’t work in that space. I do know that there are others who are already using DLT to storage some DID data though, such as iRespond.
Are there technical and ethical reasons why DLT for DID in this context doesn’t make sense at the moment? I’ll defer to those who work with those population groups. Taking all of your reasons for why it is not a good idea at face value, that still doesn’t mean that there will not be future innovations that would make DLT as a component of DIDs in this context feasible.
I am a strong proponent of responsible use of tech in the development and humanitarian sectors. I also do think that we should be concerned with inappropriate uses of technology simply because something is in trend or from complete wastes of money that really do recreate the wheel (like building bespoke solutions when existing solutions could be used just as well or better). At the same time–and call me old fashioned if you’d like–I also believe that if we want to credibly make these arguments to those who may be skeptical or to rah-rah tech is the answer to everything types, we should do so without hyperbole or exaggeration and with strong, accurate statements, using qualifications as necessary if we don’t know all the facts yet.
Literally everything, Josh. Perhaps it would be better to mount the horse from the other end: what about that document/program doesn’t have your hair standing on end? We could even turn our commentaries on that program/approach into an ICT Works post or two. 🙂
Sure, Benita, I’ll give you my take on it. What I see here is a local Myanmar organization that saw the issue of IDPs lacking access to formal education and decided to go this path. Were they guided by the lessons learned by many others before them or were they completely oblivious to that history? I cannot say. If it was the latter, should the donor and/or their implementing partner made sure that they were aware? Sure, but I can’t say that didn’t happen.
Is there anything wrong with this local organization repackaging an idea that has been done elsewhere, if it worked? Not at all. Is it wrong for the donor to hype this up as some major innovation? Certainly, that would be inaccurate and misinformed, although I didn’t really see that in my now two reads of the document. To me it was a standard: here’s what we did and what we learned document.
Now, one thing did catch my attention: the fact that they only reached 57 people. Even that though I cannot fully analyze without additional data, which I don’t have. Maybe it was a tiny $5,000 grant. That wouldn’t be so bad. Or maybe they got $500,000. That would be horrible.
I’d love to explore with an ICTworks post with you on the approach, although it would help if you would share with me (and the world) at least a few things about it that has you so troubled. Although please do keep in mind that I’m not defending their approach or saying it is how I would have gone about doing things. I only raised this in the context of not feeling that this is a good example of a development org overhyping old tech as something innovative. I’ve seen much more egregious and obvious offenders in my time.
Hi Josh – great, let’s look at doing a post on this! In a nutshell, it didn’t work, and this was predictable based on what we already know (this was one of the points that Nadia raised in her post). However, I do think that delving into the weeds of this particular example is a topic for a separate post or two. For this post that we’re discussing – the example that Nadia took is fine and it illustrates her point well. It’s not about finding the worst example that any of us have ever seen – she found a recent example that supports her argument.
Agree to disagree on this one!
Not to belabor the point here, Benita, but Nadia cited the FCDO example not to make the case that it didn’t work, but rather that it was predictable that it DID work. She took issue instead with that fact that it was being touted as an emerging technology, which I pointed out was not the case in the document she shared.
I would love to learn more about what you think that this intervention didn’t work and why that was predictable, since it is the exact opposite point that Nadia seems to be making in that part of the blog. Anyway, have a nice weekend. 🙂
@josh – the superb rhetoric you display and that I regret not to possess allow you to escape the questions: do you know of factual elements that validates the use of DLT for DID of refugees? Or not?
Taking the risk to stand by my point, I do believe that we are accountable to call evil when we see it.
With that I mind, also noting you care for responsible use of tech (and despite the fact you ditch a reasonable list of arguments I made) and remembering it takes only one to call the king is naked, let’s agree that putting biometrics on DLT for refugees is a bad idea, will you?
Nice week end to you too.
@Vincent, I’m not escaping the question. I gave you an example of one organization that is using DLT to store elements of biometric data, which has been used with refugees. Look into their solution a bit more. They are not storing biometric data on DLT. From what I remember, they store a unique encrypted key that can their hardware can decrypt. Personal data is stored separately, not on the DLT (again, from what I remember). But again, I’m not advocating for their solution.
I agree with you that we are accountable to call evil when we see it. I also did not ditch your list of arguments. My apologies if I gave you that impression. I said that I took them all at face value. That is so say, I recognize that each reason you listed are valid concerns. Having said that, I’m not willing to say there is never any situation where DLT may play a valid role in DID and that it can never be done in a way that is not detrimental. At the same time, you’re also not going to find me raving about how DLT is THE ANSWER. As with any technology, we need to seriously assess the risks of its use in advance of its deployment. If there’s a risk to harming the individuals we are working with, particularly if it is greater than the risks from alternative approaches, then of course, we should not move forward.
I’d like to wrap up on a point of agreement though. I agree that the risks of storing personal data in any system, DLT or otherwise, particularly of vulnerable populations are inherently high and so we must proceed with caution before we decide that it is really the most secure option. We must also never be blinded by excitement over shiny new things to the point that we ignore the real technical issues that you’ve flagged that may in many/most/all circumstances in the present (and potentially the future) make such a technology inferior to a different approach.
It has been a pleasure having this discussion with you. I appreciate you sharing your perspectives and pushing me to answer things that you felt I was not adequately addressing. That was the impetus of my original post here as well, to push the author to consider things that I felt were not adequately addressed.
Hey Josh – I would interpret it differently, but I think we’ve established that. As I mentioned, happy to blog this out with you at some point. In meantime, I hope you also have a good weekend – cheers!