Adaptive learning

Use Case: Adaptive learning

Classroom

What’s the issue?

With the increasing use of e-learning and blended learning, there is a growing tension between the potentially interactive physical classroom experience and the largely static, homogeneous content provided through systems such as virtual learning environments. In addition, increasing student numbers, with widening diversity, pose challenges for teaching staff attempting to provide a differentiated, contextualised and personalised learning experience to a large audience.

How can intelligent campus help?

A combination of mobile devices and classroom technology, learner data and institutional systems can integrate to understand learners’ needs and assist them to effective outcomes. A multitude of data about learners already exists, not just their activity within the VLE, but their location on campus (or outside), interactions with the library, demographic data from student information systems, their historical patterns of learning and even their ambitions or goals for the future.

Individuals’ mobile devices along with classroom technology could interact with this wealth of data available about the student and their learning to provide an enhanced experience. Personal devices and apps may also hold information on individual preferences and learning styles to enable differentiated interactions.

What could be done?

  • Digital textbooks providing hints, explanations and practice questions, linked to the class content on the VLE. Monitoring common problems could feed into course design.
  • Headsets supporting virtual or augmented reality views of content and places around the lab, campus or at sites of educational interest.
  • Adaptive devices providing personalised interfaces (including accessible tools) to meet learning preferences, and by monitoring progress adding new or supplementary content to match the needs for individual pace and style.
  • Virtual field trips (or augmented reality physical field trips) including location-awareness to give students prompts and contextual content, allowing personalised interaction with the environment.
  • Science experiments conducted remotely through the internet, with devices connected to laboratory equipment to take measurements and set analysis tasks.

Other applications use real time natural language processing to spot intentions or emotions in conversations with peers or tutors, for example using chat-based apps on smartphones. Intelligent “conversational” agents can provide tailored feedback to students on performance or respond to common questions with personalised information understanding their context, including their geo-location on campus and proximity to learning resources and facilities.

Are there any current examples?

There are few examples in practice of IoT devices integrating into adaptive learning approaches. In most cases, where adaptive learning is being used, it relies solely on institutional systems and learner analytics. These could be adapted to add integration with mobile devices.

Bolton College is providing a personalised pathway through VLE content by differentiating students according the data about them such as how they performed in previous tutorials, leading to more stretch and challenge if they do well. The college have also implemented an online digital assistant that can respond to natural language questions, in some cases removing the need for the student to access other systems such as the VLE.

At Edinburgh University, a tool using algorithmic machine intelligence has been used and quizzes designed linking to learning pathways. Elsewhere they have piloted use of a twitterbot to answer simple questions for example about deadlines. Cluster analysis and segmentation has also been applied on a wide range of historical data to identify learning trajectories and map these to current students.

coffee

What about ethical and other issues?

Security of data is crucial, with these applications involving wide ranging data including personal circumstances of students and, potentially, financial data. For example a student may use a chatbot to query how much money they are still owing on their course.

Algorithms may be making decisions and providing feedback on behalf of tutors, including on sensitive topics. Even if teaching staff have been involved in programming the behaviours and responses, sometimes an automated system may not produce an output that a human would agree was “appropriate”. It is likely that careful moderation would be needed, at least until the technology was sufficiently mature.

Who needs to be involved?

Having teachers involved in the development alongside data and systems specialists allows for a more integrated approach utilising the technology within the pedagogic context. Skill sets including instructional designers and algorithm specialists may be needed. Having a robust policy at an institutional level will help manage the ethical issues more effectively.

Adapting to needs

Use Case: Adapting to needs – intelligent accessibility

stairs

What’s the issue?

The number of students entering university who have a known disability is increasing, (56% between 2011 and 2015) and at the same time funding has faced cuts, with an impact on outcomes.

Increasing demand for services and pressure on resources suggests universities could turn to intelligent solutions across campus to improve the student experience, encourage participation and improve outcomes.

What specific challenges are there?

Going to university can prove to be a daunting experience, from busy open days and freshers week to daily negotiation of timetabling and the campus.

Universities can be physically and organisationally complex, with a myriad of different types of buildings and access requirements, spread across large areas, sometimes involving multiple campuses. Getting around campus can be a challenge and that is before any adaptation for learning is considered.

Other concerns identified by prospective students include making friends and having the right equipment.

What solutions exist?

Many “smart” devices have emerged around independent living and home automation. These include controlling your heating, security and lighting through your smartphone, sometimes to a personalised schedule or in response to location. Lighting can be adjusted to improve contrast for improved visibility, or visual signals can indicate visitors at the door, or appliances in operation. Research on healthcare and society also is developing wearable devices to support different needs.

In addition, many commonplace smartphone apps can support communication, networking and socialising. Mapping and route finding is another application readily available, such as identifying routes for public transport, cycling or pedestrians.

Other apps available include those for communicating via sign language, speech generation from text, screen magnifiers using the phone’s camera, and tasks and visual prompting tools to help with organising and attention. The intelligent campus can provide ways of integrating these personal tools with physical facilities, IT and learning systems.

campus

How can these be used on campus?

The intelligent campus is well placed to offer personalised services and adaptive facilities. Smartphones can operate apps appropriate to the individual’s needs, using interfaces including speech and natural language recognition. These could communicate with the university infrastructure, identifying particular needs within a building or area, and adjusting the environment to suit.

Navigating your way around campus might include apps for accessible routing, tailored for different needs, avoiding more challenging obstacles and estimating travel time. Real time notification of appropriate nearby facilities such as disabled toilets, parking bays, transport or access ramps could be transmitted to students as they move around the campus. Notification of special access or other needs might also be received by timetabling services to allow provision for flexible start times or remote access.

In the classroom

Example of adaptive learning environments could include

  • a lecture theatre adopting different lighting settings or activating special access features, triggered by the arrival of a student triggers (either automated or under the control of the student’s phone)
  • automatically storing content from the lecture, shared and translated into multiple formats, available on platforms that are accessed by a variety of different devices
  • automatically incorporating personal profiles of learning styles, preferences and special considerations into teaching and learning activities
  • integrating existing interface tools and apps such as speech-text conversion, captioning and gestures with learning content
  • connecting accessibility apps with classroom facilities for example to respond to head movements or translate impaired speech
  • structuring the delivery of content to allow review and reflection at the student’s own pace and comfort level

What about ethical issues?

Confidentiality and security are clearly important, with information about disabilities and special needs as well as access to specific support services and financial support. This could be even more detailed than other contexts, to include daily activities and living patterns. Students need to be confident that their information is kept private and shared appropriately with service providers. For universities there may be challenges in deciding on priorities and technical implementation of a wide range of devices and adjustments.

Who needs to be involved?

A wide range of services across the institution have an interest in such services, from housing to finance, estates to learning and teaching. Student records, IT and libraries can will also play a crucial role, alongside specialist support services for accessibility.

Entering Alpha

campus-2156430_1920

Following a successful transition meeting yesterday. The Intelligent Campus project has moved from Discovery to Alpha.

In the discovery phase we explore new ideas and emerging technology to establish which ideas meet Jisc member’s needs, are technically feasible, fit Jisc’s remit and stand a chance of becoming sustainable services. If an idea passes all these criteria, we move to the alpha phase.

The Intelligent Campus project has been looking at use cases following the co-design process in which the three ideas we came up with were use-cases which would be enabled if we could develop the basic data infrastructure for the intelligent campus.

However that is a big, long-term development so we decided that our immediate goal should be to analyse in more depth the potential use-cases as well as the technical, ethical and business implications of this approach. As part of this process we published a draft guide to the Intelligent Campus.

Having done this it gave us a better idea of where best to get started on developing the data infrastructure for the intelligent campus.

So now we enter the alpha phase. Here we will try out the most promising ideas with a small number of organisations to see what works and whether they offer real world benefits. If they do, we move onto the beta phase.

More information on how we innovate.

innovation-pipeline-november-2016

“If the walls could talk”

These are my slides from my session at the ALT Conference in Liverpool.

My twenty minute session I introduced the delegates to the concept of the intelligent learning space, one that could learn from the experiences of staff and learners in that space.

If the spaces we use for teaching and learning could speak to us, what would they say? The places and spaces across colleges and universities are some of our biggest investments. But are we using them effectively to enhance and enrich the learning journey? Does the environment in which we learn have impact on the learning journey?

There is an institutional memory within those walls that is inaccessible and lost every time the learners and teachers leave the room. The room doesn’t remember what worked well or what could have been better. The spaces, if they could store experiences and feedback, would know what worked well, and what didn’t, for different learning activities. What if, we could we use data gathered from teachers and students, as well as space usage, to inform and improve teaching and learning?

The hyperbole around AR, VR, artificial intelligence and the internet of things as created a cynical bubble among some staff and institutional decision makers, especially those that have been burned by previous tech fads. But it may be time to put aside the cynicism that this kind of hype generates and look seriously at how we can take advantage of these emerging technologies to improve the student experience, research and the management of our campuses (Clay 2017).

I also covered some of the more “creepy” elements that often arise when you start talking about gathering data about people (as well as spaces). I spoke about the importance of consent and openness when it comes to data tracking.

I presented six scenarios;

  • using environmental data to improve learning
  • using historical activity data to inform teaching practice
  • delivering appropriate interventions based on analysis of existing experiences
  • using the Internet of Things to inform teaching and learning practice
  • using artificial intelligence to plan lessons and room layouts
  • using artificial intelligence tools to influence and support learning, teaching and assessment

However in a twenty minute session we didn’t really have the time and space to discuss these in detail, we may explore these in more depth on the blog at a later date.

There were some interesting useful comments from the room (and on the Twitter) and these were captured in a Storify.

Overall an interesting and useful session.

References

Sclater, N. (2017). Code of practice for learning analytics | Jisc. [online] Jisc. Available at: https://www.jisc.ac.uk/guides/code-of-practice-for-learning-analytics [Accessed 20 Mar. 2017].

Clay, J. (2017). Could we use artificial intelligence to help manage learning spaces and improve teaching and learning? | Intelligent campus. [online] Intelligentcampus.jiscinvolve.org. Available at: https://intelligentcampus.jiscinvolve.org/wp/2016/11/07/could-we-use-artificial-intelligence-to-help-manage-learning-spaces-and-improve-teaching-and-learning/ [Accessed 20 Mar. 2017].

Amplification of the conference session at #altc

At the Association of Learning Technology annual conference in Liverpool I delivered a twenty minutes session on the Intelligent Campus. These are the responses from the delegates who attended from the Twitter.+

At the Association of Learning Technology annual conference in Liverpool I delivered a twenty minutes session on the Intelligent Campus. These are the responses from the delegates who attended from the Twitter.

If the walls could talk #altc

This year at the Association for Learning Technology (ALT) Conference in Liverpool I am presenting a twenty minute session on the Intelligent Campus.

If the walls could talk #altc

If the spaces we use for teaching and learning could speak to us, what would they say?  The places and spaces across colleges and universities are some of our biggest investments. But are we using them effectively to enhance and enrich the learning journey? Does the environment in which we learn have impact on the learning journey?

There is an institutional memory within those walls that is inaccessible and lost every time the learners and teachers leave the room. The room doesn’t remember what worked well or what could have been better. The spaces, if they could store experiences and feedback, would know what worked well, and what didn’t, for different learning activities. What if, we could we use data gathered from teachers and students, as well as space usage, to inform and improve teaching and learning?

The hyperbole around AR, VR, artificial intelligence and the internet of things as created a cynical bubble among some staff and institutional decision makers, especially those that have been burned by previous tech fads. But it may be time to put aside the cynicism that this kind of hype generates and look seriously at how we can take advantage of these emerging technologies to improve the student experience, research and the management of our campuses (Clay 2017)

If the walls, our learning and teaching spaces, could talk, what could they tell us, and how would it change what we do?

This 20 minute interactive discussion, will challenge delegates to reflect and discuss a series of short scenarios on how data gathering, analytics and appropriate technological interventions could be used to enhance and improve teaching and learning. Participants will be asked if this approach of using data gathering, analytics, the Internet of Things and artificial intelligence be used to enable innovation and effectively drive and manage change. In a similar vein to learning analytics (Sclater 2017) what are the ethical questions we need to ask and answer when looking at learning space analytics.

Delegates will go away with an understanding of some of the issues that arise once we start making our learning spaces “smart” and “intelligent”.

The session is on Wednesday 6th September at 13:00 in the Elizabeth Gallery II (moved to larger room) Library Room

References

Sclater, N. (2017). Code of practice for learning analytics | Jisc. [online] Jisc. Available at: https://www.jisc.ac.uk/guides/code-of-practice-for-learning-analytics [Accessed 20 Mar. 2017].

Clay, J. (2017). Could we use artificial intelligence to help manage learning spaces and improve teaching and learning? | Intelligent campus. [online] Intelligentcampus.jiscinvolve.org. Available at: https://intelligentcampus.jiscinvolve.org/wp/2016/11/07/could-we-use-artificial-intelligence-to-help-manage-learning-spaces-and-improve-teaching-and-learning/ [Accessed 20 Mar. 2017].

Do you need help?

coffee-2351436_1920

Use Case: Emotional Recognition in the Library

Could we use emotional recognition technology to discover when students are “troubled” or “in need of help” and then make appropriate interventions to support them in their studies? With constant emotional recognition could we start to understand what a “troubled” user’s expression is.

What’s the issue?

Users of the library often need help and support when attempting to find resources, use resources or other aspects of their learning. Users may not know where to find resources, or even what resources may be useful for them at that point in time. They may be using resources, but unsure if they are appropriate or relevant. Students may be using the library as a study space and are having other trouble with aspects of their studies.

What are the current solutions?

Most solutions rely on the students seeking support or help from dedicated staff within the library, The implication is they know they can seek support and they can find a relevant member of staff to ask. There is also the assumption that the student knows they need help, they may not see that as a possible option.

Another solution is for library staff to “discover” which students are having issues and trying to guess which students need support.

Emotional recognition software could constantly scan users and using appropriate algorithms flag students that may need support and direct staff to them as quickly as possible.

Combined with facial recognition, details about the student could be provided to the staff to enable them to tailor and personalise the support the student needs.

What about ethical and other issues?

You can imagine the strong feelings that such a system could evoke, with the idea that users of the library are being constantly scanned and monitored. Consent to be part of a system would be critical, though often consent is not sought for non-technological solutions. Another challenge would be the concern that such scanning and data gathering may be used for other purposes outside support in the library.

Are there any current examples?

Emotion recognition and emotion analysis are being studied by companies and universities around the world, most are looking at the technology to improve marketing and effectiveness of video content.

books-1835753_1920

So how could it work?

Humans can already recognise emotions in variable abilities, any kind of system to scan and recognise emotion using technology could inform and advise staff faster and quicker. Enabling a rapid response to potential difficulties.

Software already exists that utilises artificial intelligence to predict “attitudes and actions based on facial expressions”.

The challenge will be ensuring staff have the appropriate interventions to hand having identified those students who may be have problems.

There’s no room!

Use Case: There’s no room!

This use case builds on two previous blog posts I’m sorry that’s my seat! and Is there any space in the library?

library02

What’s the issue?

Demand for library space from students varies over time, not just over the day, but over the week and over the year. It can be challenging for libraries to ensure that there is space for studying. The variance in demand means that some times there is ample space and others students can be disappointed. If a student has made a special trip and can’t find somewhere to study then they may think negatively about the institution.

Another issue is reliance on group work and small groups of students having the space to study and work together in the library

What are the current solutions?

Currently libraries will use manual and automated methods to measure utilisation. Often this data is used to for self-assessment reports and proposals for expansion. Few are analysing that data in real time and presenting the information to students. Historical usage data is often accessible either through Google or Library web pages.

What about ethical and other issues?

We know that measuring usage of desks can be fraught with ethical concerns, a recent university used a sensor under desks to measure occupancy and despite efforts to inform students (and staff) the message was missed by many. It is critical when measuring usage that the library is transparent about what it is doing, how it is doing it and why.

We also know that many college libraries in comparison to university libraries are small and the benefit for any such system would be marginal.

Universities and colleges may want to consider not just demand for study space in the library, but the use of computer labs and drop in study areas.

The increase in use of digital and online collections can reduce the need to use the library as a studying space, universities and colleges may want to reflect on creating or modifying existing spaces for self study activities.

Are there any current examples

As outlined in my previous blog posts on this subject, many university libraries are already publishing occupancy in real time. Likewise Google can also provide historical usage data, but for many libraries this can be inaccurate.

Google was also providing similar information

So how could it work?

We can imagine an intelligent library which not only knows what seats and PCs are free, but can learn from history and predict when the library will be busy and when it will be emptier.

The library then provides this information to students via an app, pushing the library when there is more availability of places and computers.

Students could be notified when space is available, or for groups when group space is available.

Students could be informed of other suitable spaces in other libraries or learning spaces.

Who needs to be involved?

This kind of system requires involvement not just from the library staff, but also the team (if there is one) responsible for any kind of student app. It may make sense to incorporate the functionality required into a single app, instead of students needing multiple apps.

Ensuring wellbeing

Use Case: Student Wellbeing on an Intelligent Campus

running

What’s the issue?

As well as enabling a student to achieve their academic goals, colleges and universities have a responsibility for a student’s wellbeing. Their health, both physical and mental, security and safety, are all important to institutions. The development of intelligent campuses can offer considerable support for student wellbeing.

Safety and security

The intelligent campus is a safer and more secure campus. For examples intelligent street lighting is being developed as part of Future City Glasgow initiative. Street lighting will dim during quiet periods but the use of sensors to monitor movement and noise can bring up the lighting. The lighting could even be controlled by the approach of students via a signal from their mobile device.

rain-1599790_1920

A range of apps are already available to help students to be safe at night. A typical example is the  Circle of Six app which keeps an individual in touch with their 6 most trusted contacts making it very easy to send help messages with their location coordinates. Apps could be integrated into the intelligent campus to provide help beyond their friends when a student is lost, worried, in trouble or in a dangerous situation. An example is GuardianSentral which works with campus security and safety systems such as emergency phones and campus police. It provides GPS tracking of individuals who are in danger. It can provide check-in facilities and an arrival time at a destination. A security alert is sent if the student does not check-in. If the student is feeling uneasy the app can ask campus police to monitor them via GPS or send an officer to escort them. If they are in immediate danger there is an urgent, call assistance button which also provides the police with their current location.

Linked to a student’s safety is drinking responsibly. It’s already possible to monitor alcohol drinking using apps like WiseDrinking which monitors an individual’s drinking, based on their size, gender, time of last meal etc. provides advice on unsafe drinking and helps with calling a taxi or using public transport. In future, on the intelligent campus a student could allow their drinking in campus bars to be monitored to allow student welfare services to be alerted when drinking is excessive.

Keeping healthy

Similarly the intelligent campus could provide students with help in eating healthily on a budget, providing healthy recipes and shopping advice. The Internet of Things could monitor fridge contents and food which is out of date.

There are a large number of mental health apps now available, many of which could help with student wellbeing. The University of Edinburgh student counselling service provides a useful review of a number of these apps. They look at issues such as panic attacks, wellbeing, mental health, anxiety and mood improvement. Coventry University’s CU Health and Wellbeing app “provides information to Coventry University Students on student support services, interactive maps to NHS providers and campus buildings, and emergency contacts and help.” While this app is not really taking advantage of intelligent campus developments it does demonstrate that universities and colleges are keen to improve the health and wellbeing of their students.

In future the intelligent campus could move beyond offering advice to monitoring student behaviour and intervening if necessary. If their typical behaviour in terms of alcohol consumption, coffee drinking, movement patterns, absence from lectures, missing meetings, etc. changes, alerts and interventions could be planned by tutors or student welfare services.

Keeping Fit

Many students and staff are now using wearable fitness trackers. These could be integrated with the intelligent campus, for example, providing walking routes around the campus to achieve activity targets. Campus communities of activity tracker users could be supported creating social opportunities and even a level of competition. The university or colleges sports facilities, programmes and courses could also be integrated with the individual’s fitness targets and goals providing further support. If confidentiality issues are not considered a problem, wearable fitness trackers could provide the intelligent campus with huge quantities of data about students. This could include their sleep patterns, or lack of sleep, since many activity trackers have basic sleep tracking facilities, others can provide heart rate data which could be used to provide information about individuals with high stress levels or other health issues.

Going mobile

Use Case: Mobile Applications and the Intelligent Campus

mobile phone

What’s the issue?

Many universities and colleges are already providing student applications for smartphones and tablets to enhance their campus and learning experience. It is in fact surprising that all institutions are not offering apps, even if there is currently very limited data available for their campus.

What’s currently going on?

University and college apps at the moment typically provide information and services such as:

  • Manage library account, catalogues and e-publications
  • Viewing personalised timetables
  • University and campus news
  • Searchable and interactive campus maps
  • Find available PC’s and workstations
  • Make payments

An extensive list of features that combines a range of apps being offered by UNiversities and colleges is provided below.

However, the further integration of new apps with the intelligent campus could enhance their functionality enormously.

iphone-410324_1920

What developments are taking place?

The Smartphone developments will drive student apps and the intelligent campus can integrate with and take advantage of these developments.

For example, relatively simple campus developments can enhance the student experience, these include incorporating wireless charging facilities into desks, tables and benches. Also, provision of display areas where students gather socially, or for group work, for smartphones with built-in projectors.

The availability of augmented reality (AR) is likely to become widespread. Universities and colleges will be able to make data generated by the intelligent campus available, overlaying it onto what their smartphone camera is seeing. Real time images of buildings will be tagged with the location of, for example: their next lecture or tutorial, routes avoiding steps, location of toilets, showers, cycle racks or ATMs.

bicycles

As AR develops it is likely to become very familiar in enhancing learning. It could have a place in many disciplines, providing simulation environments for the teaching of surgery or archaeology, allowing architecture students to walk around experimental buildings or positioning historians in historical events.

Its predicted that wearable devices will become widely available, and popular, in the next few years, this will present further opportunities for the intelligent campus. Headsets, glasses, watches and wristbands will talk to the campus and provide information in new ways to the user.

Beyond wearable devices developers are discussing internet enabled implants. Futurists, such as Raymond Kurzweil, predicts that within 20 years implants (removable or permanent) will be available that will communicate wirelessly with the network. The possibilities for integration with the intelligent campus are wide ranging. These devices might record everything that the “connected” student sees and hears, indeed everything they are learning.

So what are the issues?

Clearly some of the ideas discussed here are still in development and may not come to fruition. If they do there will be a number of other issues to address. The well-known Issues concerning data security and management will need to be dealt with.

We are only at the start of the uptake of AR and wear devices which will throw up new ethical questions. Implanted devices will of course be a whole new ball game ethically and development could be halted by the ethical dimension.

For those developments that do look promising perhaps the biggest barrier is not the technological development or security questions but the need to bring together a wide range of stakeholders. In developing these mobile applications almost every component of the university or college structure, as well as local transport authorities, will need to provide data in a standard format and to a very high standard. This will require considerable resource, goodwill and management skill.

Student apps that are currently provided by universities include the following:

  • Manage library account, catalogues and e-publications
  • Viewing personalised timetables of courses, events and other activities
  • University and campus news channels and news flashes
  • Searchable and interactive campus maps
  • Exam timetables
  • Find available PC’s and workstations
  • Buy and use print credits
  • Pay for services, food and products
  • Contact tutors and other staff
  • Gym facilities availability and classes
  • Self-guided tours including videos, photos and interactive panoramas
  • Search for friends and colleagues
  • Call or email contacts
  • Push notifications of alerts and announcements
  • Alerts for upcoming deadlines
  • Attendance registration by scanning the QR codes
  • Access to cloud based file store
  • Lecture capture
  • Weather information
  • Revision resources and guidance
  • Real time cycling, train and bus information for public transport authorities
  • Details of union clubs and societies
  • Help desk and helpline access
  • Task tracking and notification – assignment and course progress.

The list of features above is based upon those offered by the student apps provided by:

  • UCLA
  • Newcastle University
  • UCL
  • Southampton University
  • Nottingham Trent University
  • Strathclyde University