Use Case: Building analytics
What’s the issue?
Universities have large numbers of buildings of varying ages and conditions, spread over a wide geographic area, sometimes multiple campuses. Managing energy, waste and resources efficiently is important for a number of reasons, including financial pressures, environmental principles and regulations, and improved working conditions for campus users.
What could be done?
A wide variety of environmental conditions can be monitored by sensors connected via the IoT to central control systems and mobile devices. These include temperature, lighting, noise and air quality. In addition, use of other resources such as water, waste and recycling
Collecting data from buildings on the current conditions and energy usage, systems can analyse needs and adjust based on a variety of factors including:
- the occupancy of buildings and rooms
- eg turning off heating and lighting when not in use or adjusting to account for the numbers in the room
- the weather conditions and use of the room
- perhaps dynamic temperature that takes account of people coming in from the cold and then warming up as they adjust, or getting colder as they sit still
- predicting the conditions for the following day
- setting heating and lighting accordingly
- informing room users of the conditions, or even what part of the room is warmer or colder
- the real time cost of energy supply
- taking advantage of cheaper energy at different times
- feedback from people currently within the building
- individually reporting their comfort level or reporting problems via mobile devices
- sensors reporting waste levels from bins to optimise collection
Another key issue is maintenance – that by understanding a more comprehensive picture of usage, the systems and infrastructure that provide heating, lighting etc can be optimally used to minimise or even predict failures and breakdowns and to provide maintenance more efficiently. Staff responsible for such systems can receive data in real time on mobile devices, respond to warnings and alerts, and more quickly be on hand to provide assistance. Routine maintenance can be prioritised to best avoid future problems by detecting the current state of systems.
Building analytics also has the potential to provide reductions in time and effort spent on building maintenance and the associated processes with live reporting and alerts.
What examples are there?
At Georgia Tech University, over 400 smart meters monitor 200 buildings. Their Smart Energy Campus Program works as as a “living laboratory” collecting data from energy utility systems across campus. They can quickly identify areas of unusually high energy usage, and provide a response, reducing time and resources. A visualisation of the campus energy usage gives them a view of consumption via a dashboard. Using thermal network and electric grid modeling, researchers aim to better understand energy usage as well as assess potential upgrades to energy systems and technology.
In Ontario, Canada, McMaster University is collecting real-time data from 60 campus buildings to optimise energy consumption. They also use dynamic-pricing data to forecast and simulate future usage.
Bristol University’s diverse buildings range from the 17th to 21st century and they use a building management system that includes alerts to the security office out of hours to monitor critical systems. A particular goal was to reduce carbon emissions, but the systems also include fire monitoring, and sustainability monitoring.
Looking further to the future
A variety of potential applications arise that are beginning to be explored in research using the data being collected from building analytics. These include:
- predicting future energy consumption including the impact of increasing battery recharging such as mobile devices and electric vehicles and the provision of charging points
- changes in energy sources such as the campus generating it’s own electricity from solar, and potentially selling it back to the national grid
- the impact of weather patterns on the availability of different energy resources and smoothing demand
- taking advantage of price variation of energy at different times
Scientists at Newcastle University’s Smart Grid Lab simulate power distribution for future scenarios. These include a severe weather power cut and increasing numbers of electric vehicles. Concepts they explore in the living laboratory including real-time notifications to users of consumption, peak demands and pricing, and problems within energy supply. They are developing a model of a “micro-grid” that provides local resilience to wider disruption.
What about ethical and other issues?
Environmental data typically doesn’t include personal data, although individual feedback on conditions could be identifiable. If health-related conditions or special needs were to be taken into account, to adjust environmental conditions, then this data is more sensitive and need appropriate management. There is a wider issue regarding the challenges of responding to individual preferences and the difficulty of getting the balance right across a group of users. Another difficulty is the lack of standards across legacy sensors and the huge challenge of maintaining and upgrading many hundreds of sensors across campus.
Who needs to be involved?
For most of the applications mentioned, the estates and facilities services are the most heavily involved, with appropriate collaboration with IT and networking services. In some examples data from timetabling and other student support services may be necessary to offer the more integrated solutions, and external data from weather forecasting or energy usage more widely.