So what do we mean by a learning space and how is an intelligent learning space different? What is a smart learning space?

As we design learning spaces, we can add sensors and mechanisms to collect data on the use of those learning spaces. It then how we analyse and use that data that allows those spaces to be initially smart and then intelligent.

Generally most learning spaces are static spaces designed to allow for particular kinds of learning. Some have an element of flexibility allowing for different kinds of learning activity within the same space.

We have seen lecture theatres where the seats can swivel to allow for discussion and group work. There are other lecture spaces where the students are seated in groups around a table, allowing them to see the front of the room and work together. New active learning spaces allow students to work independently or in groups, but the use of large screens on the table allows for whole group teaching or lecturing.

Often the pedagogy is shoe-horned into the space that is available and even if more appropriate spaces are available on campus, often they are unavailable for that particular slot or cohort.

In the past room utilisation was often a combination of what was in the timetable and what could be seen during a survey (often with a clipboard).

There is some technology already in place which can start us on the road to making better-informed decisions about how best to use space – sensors, for example. We all know when lighting is linked to a movement sensor because everything goes dark when we sit still for too long, promoting much frantic arm-waving to turn the lights back on.

But a smart learning space goes further than such simple actions and allows us to gather data about the spaces and, importantly, act on that data. We can turn down heating in rooms which aren’t being used, and some systems will take into account the external temperature, humidity and pollution levels, and not just the time of year.

We can use electronic entry systems, such as swipe cards, to ensure the security of the rooms, but also to measure room occupancy. We can also ensure that the lighting, heating and CO2 levels are within defined parameters.

If you then throw in data from the timetabling system, the curriculum, lesson planning, teacher commentary and feedback, student feedback. You then start to get a wealth of data that could be analysed and used to design and enhance the learning activities which will take place in that learning space.

A smart learning space would taken into account historical usage of the room and how people felt that the space either contributed or hindered the learning taking place there. You can imagine how users of the room could add to a dataset about the activities taking place in the room and how they felt it went.

Of course there is a challenge with historical data in terms of bias, errors and legacy processes. You can imagine that if a space, regardless of what it had been designed to be, was only used for lectures, then the historical data would imply that the space was only ideal for lectures. Bringing in more datasets would help alleviate that issue and ensure any assumptions about the space had some element of validity.

You would think that data from the timetable could allow for this automatically, but timetabling data tells us about the cohort, the course they are on and the academic leading the session, most timetabling software doesn’t have the granular activity data in it. What will be happening in that session, not only what was planned, but also what actually did happen.

The course module information may have the plans of the activity data within it, but may not have the room data from the timetable, nor may it have cohort details. You could easily imagine that some cohorts may be quite happy with undertaking group activities in a lecture theatre space, but there may be other cohorts of students who would work more effectively if the space was better at facilitating the proposed learning activity.

Likewise when it comes to adding feedback about the session, where does that live? What dataset contains that data?

Then there are environmental conditions such as heat, temperature, humidity, CO2 levels, which can also impact on the learning process.

So an actual smart learning space would be able to access data about the session from multiple sources and build a picture of what kinds of learning spaces work best for different kinds of learning activities, taking into account factors such as cohort, environmental conditions, the academic leading the session and so on…

These datasets could also be used to inform future space planning and new builds, but smart learning spaces are only the beginning. Taking a smart space and making it intelligent is an obvious next step.

An intelligent learning space would take this data, and then start to make suggestions based on the data. It would identify possible issues with the learning plan and make recommendations to either change the learning activities planned, or recommend a more appropriate space. An intelligent learning space would adjust the environmental conditions to suit the activities planned for that spaces, rather than users of the space having to manually adjust the conditions when it becomes too cold, too hot, too bright, stuffy, etc….

An intelligent learning space could take data from a range of sources, not just the physical aspects of the space and how it is being used, but also the data from digital systems such as attendance records, the virtual learning environment, the library, student records, electronic point-of-sale and online services.

This joined-up approach can provide insights into the student experience that we would otherwise miss. These insights can inform and support decision-making by individuals across the campus, including students, academic and professional service staff. By using live and dynamic data, decisions can be made that are based on the current state of the different learning spaces across the campus.

Making the timetabling software intelligent, well at least dynamic, could mean that learning spaces are not allocated to cohorts of students for a set amount of time, but learning spaces are allocated based on pedagogical need and student need and done as and when needed.

One of the key issues with all this is to collect and store the data somewhere, a centralised hub or data lake would be critical.