This week at HQ, we’ve started doing something called “deep work”. The people at Do Lectures, whose book we’re reading this month about writing good email newsletters, recommend it as a way to not be distracted by all the things that pop up in our lives now thanks to our phones and the web and all that. It’s good! We’re going to persist.
Our deep work mechanics at this stage are that I’ve set an alarm on my phone that goes off at 2pm, and again at 5pm, and in that window, we try not to use our phones and turn off the WiFi on our laptops.
One of the main jobs we have at the moment is to think more about sales and marketing. We haven’t especially done any yet, apart from talking about our work to people who mostly already know us, so we’d like to broaden our audience a bit – hence reading about writing better newsletters. Would you like to sign up for our newsletter? We thought we’d start with our homepage, which hasn’t been changed since we wrote it, really, back in 2016. (Not proud of that!). So, we’re working on that.
During that exercise, we were flung off into a lovely exercise of pulling things off our shelves, and arranging it — or knolling it — on our big work table. It was really satisfying to see how Museum in a Box has evolved.
Here’s a wobbly panorama of the whole scene to start with. Packaging, Box inserts, the Box and its various design branches, the “brain” which is all the hardware/software inside, our progress bar, and the sound elements, from volume knob selection to amplifier design.You can see the Box insert cards, from the very first idea of a scribbled “Updates” card, to our current TRY ME! cards, which give people who’ve bought a Make Your Own kit something to play before they’ve finished their collections.These are the various elements of the “brain”. From earliest at the top, to most recent at the bottom. Exciting to see Adrian’s very first comp of our physical progress bar, which we developed because the Box took about 30 seconds to start up. Now it takes about 12 seconds!And finally, we have about ten hardware versions we consider to be major milestones in the design. It’s not quite Dyson’s 5,000 iterations, but we aren’t as rich as him, so, there’s that. It’s still really satisfying to track the design evolution and how we’ve continually synthesised feedback and technical considerations.
We’ve parcelled all the iterations up into their own labelled boxes on the shelves instead of them being all over the place, which I find comforting.
As we move into 2020 we want to take a moment to talk about the latest version of the Box, version 1.3. We’ve dedicated much of the last year to developing it, and we LOVE it!
Behold version 1.3! Think it doesn’t look much different? Think again and look under the hood…
At the beginning of 2019 we sent 40 boxes around the world in our pilot. Following the feedback from that, and coupled with a wish list we had already built up, we identified a number of ways we could improve the Box.
At the same time we set ourselves the ambitious goal to build 1000 boxes by the end of the year. Building large numbers of the previous design wasn’t ever going to be practical, it was made up of a tangle of wires that we had to hand-wire ourselves and it relied on a variety of components ordered online from a multitude of suppliers. The Boxes were also subject to the occasional injury when transported around the world, so it’s suffice to say we had our work cut out!
Some of the ways we sought to improve the design included:
Better audio – louder and clearer!
Faster assembly time
Fewer parts
More durable design
Logging boops offline
And so we started work in the spring; Adrian worked on designing the PCB as a neat home for all the components that were previously crammed into the box, George developed the awesome instructional graphics on the board and worked on software improvements, and Charlie designed a new acrylic ‘skull’ and mapped out the positions for the electronics and how they would be mounted inside the Box. So here it is…
The new PCB…
We LOVE the new PCB and it has a few important features worth talking about. Importantly there’s no wires. Previous Boxes included a large micro-USB extension cable, aux jack, and loads of wires which got in the way. Now the jacks are tiny components that sit at the back of the board and are devoid of the tangled wires that criss crossed older versions. This means we no longer have to hand solder anything (Yay!) and the boxes will be far more durable, reliable, consistent.
The ‘Brain’ of the Box is now a neat stack of three separate boards: A Raspberry Pi 3A+ sits at the bottom, then there’s our main custom board and progress LED board (affectionately known as ‘Blinky Lights’) which slots in a right-angle socket; and finally we have a stacking header and plastic standoffs which raises the reader high above the main PCB just underneath the surface of the box. Primed and ready for Booping!
Our beautiful board (note the handy prompts to help you understand what’s what!)
There’s now a REAL-TIME CLOCK (RTC) too. Previously if a Box wasn’t on WiFi we had no way of logging when a Boop had happened because the Pi doesn’t keep track of time when it’s not online. So this new RTC allows us to timestamp a Boop and log it next time the Box is connected to a WiFi network.
Testing the board and speakers for the first time
Then there’s the DIGITAL-TO-ANALOG CONVERTER (D.A.C.). As the board says the D.A.C. ‘converts digital audio into analog sound for the amp’. This along with our super swanky and loud speakers make for waaaay better audio, that is free of static and incredibly clear. This alone makes the experience of using a Museum in a Box, particularly in noisy environments, so much better.
1000 custom made speakers!
The new speakers not only pack a punch but also weigh a lot less which in turn makes them far easier to mount. V1.3 is almost half the weight of previous versions which will reduce both the cost of postage and it’s environmental impact when shipped around the world.
The new Skull…
A comparison of the Box design evolution from 2018 to 2019
With our new PCB came the challenge of mounting it inside the ‘Skull’. Past versions required us to fix bits to the sides, top, and base with an opening on the underside. This was fastened by screws in the feet and some t-nuts that were an incredible pain to mount and often came loose!
MiaB team visit ProtoLabs
The advice we kept receiving when talking enclosures was to injection mould it. The main benefits of that being a moulded enclosure is ‘preassembled’ and ready to put the electronics straight in. It’s also very scalable and would allow us to integrate snap fittings into the design for mounting components inside.
To explore this we did a lot of research into moulding and a bunch of work CADing and prototyping different enclosures. We also took a team trip to visit Protolabs in Telford and learned a great deal from the brilliant people there and received an exciting factory tour of their setup. We attained quotes but ultimately decided not to go with injection moulding for two reasons:
1. The upfront tooling cost is very high and hard to justify for the small batches we planned to initially produce.
2. The overall aesthetic of the Box has become part our identity. The ability for us to easily modify the design, and teach others to build the Box (such as our amazing Verizon volunteers!) feels much more appropriate for our scale and values.
So, having decided to stick with our laser cut look we needed to find a neat way to mount the PCB inside the Skull. The solution was to switch from a bottom opening to a front loading Box. We created a neat groove on the back panel for the PCB to slide into and sit neatly in the middle of the box, lining up with openings at the back for the micro-USB and aux jacks. The Box can now be opened, the PCB removed, and replaced in mere moments!
Get in the groove! The ledge the PCB slots and rests into, lining up with the aux & power jacks on the skull.
The extra front and back panels are held in place and made removable by using some neat snap rivets which can be removed from the outside. We used these rivets for speaker mounting too where before we’d faffed about with tiny nuts and bolts and a laser-cut stand.
Blinky Lights – We use these to indicate that the Box is ‘warming up’ and ‘ready to go!’
Other new features include a new progress LED board which is slicker and uses some recycled acrylic offcuts for the shims that we sent to the PCB manufacturers European Circuits. We spent a long time looking for the perfect light pipes to use with surface mount LEDs but ultimately decided to stick with through hole LEDs and our big green green LED. Why? Because they look bright and amazing!
New Packaging!
A new Box deserves new packaging to go with it. We revisited a previous design using a more compact container and an insert that conceals the power plug and various admin and try me cards so all you see when unboxing is your shiny new Box!
The new packaging which we tested by passing the box around the office like a rugby ball!
Finally there’s the latest version of the software. We’ll talk about this in more depth another day but we’re chuffed that the Box now boots up waaay quicker than it used to and includes more audio guidance when working through steps like WiFi and updates.
The new assembly time…
With our new design manufactured we decided to assemble the first 20 Boxes. We were blown away to discover that assembly now takes ~7.5 minutes per Box – a 2000% increase from the old design!
George and Charlie building 20 boxes in 2.5 hours!
In future this saving will enable us to fulfil orders quicker than ever before. The Box was intentionally designed for disassembly and as a result we proudly no longer use glue during the assembly process. The Box will now be much easier for people to disassemble and recycle at the end-of-life and we will share a breakdown of the parts and materials inside the Box in due course to make safe and appropriate disposal even easier.
Assembling our first batch of v1.3 Boxes in HQ
And so with our shiny new Box design came our first test…
First boot up of v1.3… we figured out the Zeus print, eventually!
So there you have it, our new Box and the culmination of over nine months hard work. I hope that proves an interesting insight into some of the design decisions we made in developing v1.3. We love it and we hope you do too!
Our shop is now online so you can buy your very own v1.3 Box with a Make Your Own kit now! Kits are available in Individual, Educator, Small Org and Large Org options and colours include CMYK, Transparent, and Plywood.
Box colours include: Cyan, Yellow, Transparent, Plywood, Magenta, and Black
Have a wonderful New Year from all at Museum in a Box.
When people see and try a Museum in a Box for the first time, after they’ve been charmed by the magic and wizardry, they’ll often ask how does it work?
Once we start down that line of conversation, it also doesn’t take too long before people ask have you patented this?
The fun part is, we tell people exactly how it works. It’s a combination of Near Field Communication (NFC), a Raspberry Pi, and a few other bits and bobs, which we happily explain.
Illustration of main components, by Charlie Cattel-Killick, February 2018
We’ve also made boxes out of transparent acrylic so you can see exactly how it works. I was chuffed when I handed one of our transparent boxes to my nephew, Alex, and he was able to identify, describe, and connect each of the hardware components into the whole. And I’m not just saying that because he’s family. I was genuinely impressed.
A transparent box alongside one of our early physical prototypes for our new internal layout
In one of our recent commissions, with Monroe County History Center, we loved hearing that they thought a transparent Box in particular would be of interest to their community. The commission was about developing reminiscence collections for folks suffering from dementia and their carers, and quite a few of the men in the programme had an engineering background, so we hoped there’d be interest in the device itself, as well as the 1960s.
We also want to make our design available in an open format so students can literally make their own boxes too. I suspect we’ll end up doing this with one of our early designs which maker spaces or classrooms may have the components to hand to fabricate their own Box. As we’re marching towards higher production volumes, that’s meant developing internals that are much quicker to assemble, and can be made in large volumes. We are developing a Make Your Own kit version this year, so far focussed on the content side of things – like choosing and digitising objects, writing and recording their stories – but we’ve always hoped to extend Make Your Own to include the hardware piece too.
That brings me to our newest design innovation, which relates to our new and improved printed circuit board (PCB). Instead of us hand-crafting the internals – because I should never be making wiring looms! – we’ve now got a super-slick, all-encompassing PCB which incorporates and unifies the amplifier, digital-to-analog audio conversion, power for the Box, our physical progress bar, and allows us to add a real-time clock to help us manage and record when objects are booped.
As Adrian and I chipped away on the PCB layout, I found myself unable to recognise or interpret what all the wiring bits actually were, and needed Adrian to label them for me. Then it dawned on me. Why not label them for everyone? Why not explain transparently exactly what each bit does? So, I designed cheeky labels for each of the main components, so you can familiarise yourself with it all.
Our shiny new annotated PCB!Blinky lights!!!!
We’ve also redesigned how the Box opens up, to make it easier to pull the guts out and have a look. People – often men, interestingly – tell us “but you could just do this with a phone”, and we nod politely and say yes, you could. But, a) that’s just another screen in your life, b) you can’t just pull your phone apart to understand how it works, c) we want the Box itself to be part of the experience, and d) it is also an educational element of the whole idea.
I’m still not sure if we’ll patent the hardware design, but I at least wanted to establish a state of the art on the annotation-on-the-thing idea with this blog post. That might mean we get totally stiffed by some giant megacorp, which would obviously SUCK, but, I prefer to radiate generous intent.
The Jewish Museum London (JML) commissioned a Box from us last year, about remembering Jewish people who served in WWI and WWII.
They got in touch with us again this year to see about doing an upgrade of their Box – our first upgrade!
Since we delivered the initial commission, the Box’s core design has seen some important changes, and in fact, we declared V1.0 of the physical design back in February this year. We’ve also been gradually improving the software that runs the core Box interactions, so this was a good chance to upgrade the code on the JML Box while we were poking around.
Physical upgrades
The plywood ‘skull’ has been simplified with the speaker and power jacks now exposed on the back of the box, along with access to one of the Pi’s USB ports. (We’re not sure what we want to do with that yet. One idea is to incorporate a microphone into the mix.)
Most of the previous perforations across the older ‘skull’ have now been filled in, to focus the sound (so it doesn’t just bounce around inside the box), and most importantly,
We’ve replaced the old tiny stereo speakers with one beefy new mono one (see below)!
Our beefy new speakers!
So, to do the upgrade, we took back the museum’s old Box, gutted it, reusing what we could before cutting them a new V1.0 Box in plywood, reconstructing it with the new sound components, and handing it back.
Jewish Museum team checking out the box
The original 2017 box
The upgraded box!
There were two other important aspects of the upgrade: the audio clips and the software. The museum had tested their Box in schools and found the audio was too long. We’ve been evaluating this challenge across all the Collections we’ve made, and developing a much better understanding about audio duration and content types that work really well with young children. Our main conclusion is – perhaps unsurprisingly – if clips are too long people begin to disengage and switch off. Therefore, the museum trimmed some of their lengthier tracks down and we republished them to the Box.
Now, with WiFi!
It’s now possible to configure each Box in situ to get on to a WiFi network, so we made the cheeky addition inside the shiny new packaging – a WiFi card! The card allows the box to connect to a local WiFi network with the assistance of a smartphone, tablet or computer.
The upgrade gave us the opportunity to test the new wifi configuration out in the wild for the first time as well as update the shortened audio tracks onsite using the museum’s WiFi. Once the box was online, and after a little troubleshooting had been done, the box automatically pulled down and updated the new audio tracks!
What WiFi means for the future
Upgrading the Jewish Museum London’s box has been a great testbed for us to learn how we can retro-fit and improve upon older boxes as well as provide on-site updates to content without the need for physical intervention from us- this is an exciting development that’s heading towards our long term goal, where we can offer subscriptions to people who have Boxes. Say you like Natural History” and you subscribe, every month (or so), you get a new set of things delivered from museums all over the world, and your Box just knows about the new set of things because we’ve been able to update it in the background.
Huge thanks to the JML crew for inquiring about the upgrade and for being patient while we figured out a method of best practice!
In preparation we’ve spent a bit of time working on new packaging for what we refer to as the ‘starter box’ because it contains: the brain, an introductory print, and a mains plug; everything you need to get started! Previous versions have been bulky and wasted space which can make travelling with the box more challenging than it need be. We wrote exactly a year ago (gosh!) about our visit to MOO HQ where we worked on packaging ideas with their product design team and having now spent some time designing a new box we wanted to share a brief overview of our packaging evolution and developments:
Statues of Women in London – December 2015
December 2015: Early brain prototype combined with the statues of women around London set. We later decided to separate the brain into an independent starter box because the variation of possible 3D prints and postcards in the box is so great. This gives us more freedom now when packaging up each unique collection.
***
MOO Box – November 2016
November 2016: An experiment developed with Moo Ltd using their plotter machine and a square brain. This included a space for a set of NFC cards.
***
February 2017: Starter box comprised of many laser cut layers of card. Solid but used an unnecessary amount of material, was heavy and there was a great deal of wasted space.
An object box made for the Jewish Museum London’s ‘Hearing History’ set
***
Render of a foam insert design – July 2017
July 2017: Stacked foam insert, the brain sits above the starter introductory print and plug. We liked this foam design but for small quantities we’d need to laser cut and stack multiple layers which is both time consuming and costly.
***
Development of a card insert – October 2017
October 2017:
In this new design the brain sits above the insert, when you remove the brain the plug and print tucked away underneath are revealed.
We adapted the foam design to allow for the slightly bulkier Rasperry Pi plug. Doing this means we could easily change the plug type and send a starter box to almost anywhere in the world without the need for an adapter… it’s the small things that count!
The box is made out of three parts: the box itself (a), the insert (b) and the tray (c) glued to the underside of the insert. The real challenge was designing an insert that works with the many different plug adapters of the Raspberry Pi universal plug!
Raspberry Pi power adapter
The packaging is a great improvement on the older bulky and heavy box making it much better for sending in the post. Having travelled around with this new box however, we’re now thinking we can make it smaller still!
Hopefully this has been a brief but insightful overview for any packaging nerds out there!
We’ve made about 20 prototype boxes now and have learned a great deal from each one. We wanted to highlight one particular box we made a couple of months ago where we experimented with a smaller form and what making it has taught us.
The design of the box or ‘skull’ (the plywood/acrylic case that contains the tech) as we refer to it is dictated by two things: the form factor of the Raspberry Pi in question and all the features we feel necessary for the product to have.
Early on we were creating boxes with the Pi 2 which required a dongle to connect the box to the internet but several months ago we switched to the Pi 3 which features built-in WiFi saving space within the skull. Raspberry Pi also make the ‘Zero’ which is about half the size of the Pi 3, we liked the idea of a small box which would be more transportable and also not require mains power connection so we designed a smaller square brain inspired by the recorder box we made back in October.
Our prototype recording box which inspired the square brain design
I (Charlie) got to work with the layout of the hardware inside the box trying out a new method of speaker mount while Adrian worked his tech wizardry to figure out what hardware to adapt and then got cutting! The square brain featured several changes from the regular rectangle namely:
a power on/off button
push button volume control
No LED progress bar
an internal battery charged via a Pi charger board and micro USB cable
A single speaker mounted to one side
We tested the box at Nottingham’s Explorers Fair (we’ll share a post on that soon) where we had it set up alongside the standard rectangular box. Seeing the two side-by-side it was clear the rectangle with its larger surface area provided more of a platform for the children to place multiple objects on top of however the square allowed them to pick the box up and put it to their ear or sit down on the the floor with it.
Getting hands on with the square design at Nottingham’s Explorers Fair
Despite working well and having great mobility the square box also had some obvious limitations:
the Pi Zero only allowed us one speaker, so the sound wasn’t as good
the clicky volume buttons weren’t as effective or efficient as a dial
the lack of our physical progress bar didn’t help people understand they had to wait a bit
larger objects might not balance well on the smaller top
The square design with its illuminated power button and push button volume controls
We do love the smaller form factor but when you put the two designs side-by-side the larger rectangular box has a greater presence, not to mention more room for fiddly cables and components. It was a great thing to prototype and has since influenced alterations for our bigger boxes. This won’t be the last you see of square boxes however, I’ve had some fun recently prototyping a bigger ‘Design-for-Disassembly box, but all that is for another day.
Before I get rambling, allow me to introduce myself. My name is Charlie and I’m the latest addition to the Museum in a Box crew. I spent some time with George and Tom as a visiting researcher shortly after their transition to the new Bloomsbury HQ in the spring; but now spring has sprung and summer is done I’m lucky enough to have joined as Junior Designer and officially one of the team. On with the post!
It’s been a while since we’ve written about the various goings on at Museum in a Box so I’d like to take a moment to break down a few developments of late.
Up until now the product has revolved around two parts: the Brain (the box) and the artefacts, either 3D prints or 2D postcards. The interaction too has remained relatively unchanged with an object placed on the box triggering audio which provides context to its origins and history. The audio content that plays has often been a unique script written and narrated by one of the team, however as a great deal of our focus is on the box being used within the classroom, providing information around an object is only half the interaction. Seeing as each box is powered by a Raspberry Pi we’ve begun to explore new ways that we can use that to ramp up the fun that users can have with the box.
A box that listens?
When we consider how the box may be used in the classroom one important factor is how it can act as an intermediary between pupil and teacher. We’ve seen how well school pupils react and get stuck-in with the box and its objects, so finding a way for that to fit in with lesson plans and teaching is important. In response we’ve started work on a new component, a recorder box which will contain a microphone and aims to compliment the box and its objects.
We’ve had the idea of being able to record to the brain in our minds for some time and we’ve been asked again and again if a record feature is on the cards. Seeing as it had come up in discussion so often we thought it was time to crack on with a spot of prototyping!
Mocking up a cardboard record interface
We begun by mocking up a few cardboard interfaces with various button configurations, as well as postcards which could control what mode the box was in but ultimately decided that if the recorder box is plugged in, and an object is placed on the brain, it will switch to ‘record mode’ and announce just that. The box will then issue a count down and allow the user to record their own description of the object with the options to replay, delete and save their recording. This new recording will now play every time that object is ‘booped’ unless a newer recording is made.
A render considering what form a record box could take.
The recorder box will alter the way we think of the content associated with any given object, instead of it being a predetermined description, by plugging in the recorder the user can create their own customised audio. With the object descriptions essentially in flux the recorder will lend itself very well to classroom activities where teachers could set unique tasks to objects that pupils can then record over with their own responses.
We’re now starting to talk to teachers, and understand who the product is serving and how we see the record box working both inside and outside the classroom: whether it’s museum staff creating activities for visitors or parents recording stories for friends and family, we’re super excited to be prototyping this new tool and will keep you updated as things develop.
Over the past nine months or so, we’ve been able to show Museum in a Box to hundreds of people, either in our office or at events, and the response has been fantastic.
It’s also been ongoing informal user research, and we’ve had the chance to watch people figure out how to use it. We’ve varied our description of the mechanics and amount of setup, and observed (very casually) little sticking points. One of the main things we noticed is that it wasn’t clear when the Brain was ready to go. The Raspberry Pi 2 takes a while to start up and get ready to read an object, about 30 seconds, actually. So, we’ve added a physical progress bar to the box to help people know when it’s ready. It even says READY!
Adrian soldered the first version, which you can see here. We also adjusted the layout of the box to simplify it a bit. All you really need to know about is power, volume and when it’s ready.
And we put a BIG GREEN LIGHT at the end, which is fun.
I’ve also been thinking about what kind of simple instructions we’ll need to include in a box that doesn’t have us driving it. Hopefully something like this, with just three steps would be good.
And then, Adrian took a very exciting step and ordered us our very own Printed Circuit Board (PCB) to drive the progress bar from now on.
It’s possible I’m overexcited about the progress bar, but, I love it!
One of my favourite things about museums in boxes is that whoever is looking at the stuff can arrange them however they like.
Even in our very first version, on the very first day, all we did was arrange the stuff. It was interesting to arrange them by the date they were made, and the date they were acquired by the museum. We also mapped them geographically too.
Here’s the display we made that day:
We’re working on our first commission – Woo! – and it’s a set of Big Stuff from the British Museum. You can obviously arrange them by size, but also when they were made, and when they were acquired. (Note that the human-sized figure is the box commissioner’s wife, and not part of the British Museum’s collection.)
Fun to think about how these sorts of arrangements could be transformed into different information. At the moment, our boxes have a single point of contact, but there could be many. Maybe you can arrange the objects to get different stories.
