On 6 March, the Beneficial Artificial Intelligence Society hosted Simon Chapel at their weekly discussion to talk about student data tracking at the University of Edinburgh. Mr Chapple is the current head of data technology at the university. The first half of the night involved an explanation of what data the university can track and how it uses it, focusing on how the university tracks library capacity.
The university tracks our data by using the Internet of Things (IoT), which is an interconnected system of internet-connected devices that can sense, share, and use data. Chapple discussed the issue that the university was facing: the university library can be packed and overfilled, especially during peak hours and exam times. The fact that the library was recently made 24/7 also makes it harder to use conventional methods of tracking capacity.
Chapple and his team have been tackling this issue from various angles. The rest and perhaps most obvious is the key swipes when people enter and leave the library. This gives an exact count of how many students are in the library, as well as how long they stay. While this helps keep track of overall capacity, it can’t track what foors students are on and where free seats are. In this respect, Chapple’s team can use desktop PCs to track where students are sitting. But again, this has a major drawback in that often students sit down at desks with desktop PCs but use their own devices or books.
There are infrared sensors that can be installed under desks to effectively and accurately track student presence at study desks, allowing the possibility of setting up a system to guide students to empty desks. However, these infrared sensors cost upwards of £250,000 per year, making it economically impractical.
Chapple’s team has come up with an economical solution to the problem by using the university wi system. Every time your device connects to eduroam, it connects to an access point (AP). The university has wi APs all around campus, but the library in particular has an abundance of them to avoid black spots without wifi . When your device connects to an AP, the university gets a few important pieces of data: the fact that you have connected, which AP your device connected to, your identity, through eduroam, as well as what type of device is connected. However, Chapple assured that the university only processes the rst two pieces of metadata, discarding the others.
Because APs are often near clusters of study desks, his team can use your connection to the AP to provide information about where study desks may be available.
Currently, the university only uses this system in the library, although various university departments such as estates would like to see implementation across campus to track how students use facilities. Chapple emphasised that the university currently does not and has no plans to use more metadata than right now. However, he warned that there could be government pressure on the university to use more data tracking for various goals.
It’s an ethical dilemma to be solved; how much can the university do with our data without it being an infringement on our privacy rights? The university plans on consulting students in the future to give feedback on this data collection.
Image: Stinglemhammer via Wikimedia Commons