Museum Egg

We made a prototype of something we’ve called Museum Egg.

We built it during the Science Museum London Digital Lab Hackathon. The Hackathon was a brilliant 2 day event where the Science Museum and it’s data became our creative playground.

The idea is that the Museum Egg is carried around the museum by a visitor. It tracks the things they look at and spend time in front of. At the end of the visit, when it’s returned and docked, a personal postcard is printed of that visitor’s favourite items.

Museum egg logos

We wanted the Museum Egg to provide an experience that:

Contributes a memento

A unique physical keepsake that’s a reminder of a day at the museum

Is Personal

Each visitor is reminded of their own favourite exhibits.

Has no screens

Visitors can keep their phones in their pockets and concentrate on the exhibits on show.

Has frictionless technology

No need to download an app or rely on public wifi. No log-ins to create and no codes to enter or scan. No fears about foreign data roaming charges or phone battery drainage. No problems with permissions and privacy.

The Museum Egg prototype works in four stages:

1: Pick up

The Museum Egg turns on automatically as it’s lifted from the starting dock.
Pick up

2: Wander & wonder

As the visitor walks around the museum, the Egg receives data from bluetooth beacons. From this we can work out where the visitor lingered the longest and which exhibits they were closest to.
Exploring the museum

3: Dock

At the end of the visit the Museum egg is returned to its dock where it automatically downloads the data.
Docking at the end

4: Print out

The data is processed to find the top three dwell times and their locations. The Collection Online API is used to find details of the corresponding exhibits. A postcard is printed showing these 3 ‘favourite’ items and a fourth item that is similar but not on display. This is a secret item that’s only shown to one visitor each day.
Postcard print

Technical overview

Inside the Egg

The egg contains a Raspberry Pi Zero, Micro SD card and a Bluetooth adapter so that the Pi can scan for Bluetooth beacons. The Pi runs on a 2000mAh rechargeable battery along with an Adafruit PowerBoost 500 module to handle the battery charging and connection to the Pi. A green LED was included so we would know the device was working correctly.

Bluetooth Beacons

Bluetooth beacons are low energy devices that send out a small packet of data, usually at a one second interval. This data packet contains several configurable IDs that are unique to each beacon. Signal strengths can be used to work distances from the beacons, but it isn’t too precise. In our prototype, for simplicity and speed, we gave each of our exhibits it’s own beacon. This meant that when the Museum Egg detected a nearby beacon, we knew exactly which exhibit it was next to.

Starting and Logging Data

There’s a designated beacon in the starting dock and once the Museum Egg detects a strong enough signal strength from this, it starts logging our exhibit beacons. Some simple processing is done to work out each signal strength, which gets logged alongside beacon IDs and time detected.

Data Transfer and Processing

When the Museum Egg detects the finish beacon, it stops logging new data. It then saves the data log into a CSV file and starts a file transfer over Bluetooth. A Node JS script processes the data by creating a score for each exhibit using the amount of beacon detections over time and proximity to the exhibit.

Postcard template and Printing

The list of beacons and their scores gets passed to a second Node JS script. Using this data and a Pug template file, it includes photos and descriptions of the top 3 scoring exhibits, pulled from the Collections API. Once the template is full, it is saved as an image file and printed with a mini photo printer.

Links & resources

Our code: GitHub
Our Hackathon team: Brandon, Matt, Louie, Mike & David
Science Museum website: sciencemuseum.org.uk
Explore exhibits: Collection online
Mar Dixon’s Instagram Storify