UNSW CoDe 2016: Data Driven Workplace
End of semester 1 review
Arup in Australasia is renowned (in the Arup world anyway) for having cool offices. In particular in Sydney, our regional HQ and home base of the core digital team, we are currently trialing new ways of working before moving home in a couple of years. BVN have done a great job working with us designing the space and which has created a much more fun and collaborative vibe in the office.
As part of this process we are investigating how we can make our workplace more data driven, as WeWork notably are too, particularly as an Activity Based Working (ABW) environment is a significant shift for the majority of our firm. We have framed 3 main areas of interest:
- Allows people to find individuals within the workspace in near-realtime
- Provides individuals a more informed choice in ABW environments and encourages movement to use the spaces as they were designed
- Allows us to understand how our workspace is utilised to inform future designs
To help us research this we are pleased to be supporting the first group of honors students from the UNSW Computational Design (CoDe) course from the faculty of the built environment. Even better is that this we’re partnered with BVN on this, so the new workspace design can be measured and iterated, informed by the data we are collecting.
We have 3 students working with us, Anissa, Alex and Tiara, who have just completed the first semester of the project and last week presented us with their findings. They maintain a blog of the investigations here: http://where-in.space/
After the initial research phase the premise of the solution currently being investigated is using iBeacon technology, but almost in the reverse of what it was designed for (providing context to devices/users when the are within range of an iBeacon signal).
After the technology landscape assessment our initial approach is to deploy a mesh of base stations (raspberry Pis), and use these to scan for Bluetooth signals emitted by small iBeacon’s carried by people, allowing us to determine (or make a reasonable guess at) which base station each user is nearest.
For privacy concerns we did not want to scan for individual’s devices (laptops/phones) and we wanted a clear physical opt in process, so giving people the option of carrying a small beacon which we can scan for as opposed to tracking people is a much friendlier way of doing it.
The iBeacons themselves broadcast a data packet which we can use for a range of purposes, the main ones being UUID (identifier for a batch of beacons, used to designate all Arup beacons), a Major ID (used to assign an Arup office), a Minor ID (used within each office to assign to each user).
On top of this for each transmission the other useful one is Received Signal Strength Indicator (RSSI) which is the strength of the signal as received by the rpi (which we use to guess how far away it is).
The rpi base stations then listen for these beacons, and when one is detected a local express.js application sends this to a database — in this case MongoDB.
From here we can query this database to see where each individual beacon was last seen, and present this to users to answer use case 1: Allows people to find individuals within the workspace in near-realtime. Beacon location alone is not enough to provide a user with an accurate result so we’re looking at integrating other datasets to supplement this include port mapping from docked laptops and Lync status.
As well as the rpi listening for iBeacons we’re also using them to monitor the environmental quality of the office, testing a range of sensors including the Sense Hat. Using this data, plus the locations of colleagues and teams were hoping to answer use case 2: Provides individuals a more informed choice in ABW environments and encourages movement to use the spaces as they were designed. It will be interesting to see if providing more information on the workspace in realtime will have an effect on people’s use of it.
Once this has been up and running for a while and we have some data to analyse we can look at use case 3: Allows us to understand how our workspace is utilised to inform future designs. Not only will the data from this platform show us how environmental factors influence how people choose workstations (or not) but also hopefully help us see which office configurations are effective.
So that’s the theory and as we reach the end of semester 1 the application is getting there. We’ll post another update here once we get some results…
