An innovative feasibility study using sensor data to establish feedback loops to help improve building performance and wellbeing

Interest in Building Information Modelling (BIM) has seen a considerable rise in consistent data sharing practices, streamlining the design and construction processes.  However, owners and occupants still do not know if their building is performing as ordered.  Therefore, there remains an opportunity for work to improve the utilisation of information once a building is in use.

Buildings can generate huge volumes of data when in use, which can be collected in several ways such as sensors monitoring temperature, humidity, daylight and energy consumption.  Maintenance data (eg repair requests, service plans, product guarantees, etc) and occupancy information, such as the number of tenants, age groups and satisfaction can be collected by various means such as PROBE occupant surveys.  Data is typically spread across multiple disconnected systems and in numerous formats, which subsequently means drawing conclusions from performance gaps is limited.  Better data gathering protocols and analysis models can provide advice to building managers, to mitigate poor performance in use and close the loop to designers and construction to evidence the causes of the gaps.

A performance gap is a disparity between predicted and realised performance, for example energy use predictions in the design phase, and the actual energy use of a building when in use.  Performance gaps occur for many reasons including assumptions and lack of data which typically arise in the provision of design details to the contractors on site.  These design details are either unbuildable, loosely defined, or missing.  Improved feedback from in use phases can help close such gaps so that lessons can be applied earlier in the design and construction phases.

These gaps have greater impact in sectors such as social housing due to inaccurate assumptions about building occupants. For example, poorly designed buildings with vulnerable tenants can severely impact fuel poverty and occupant wellbeing. In 2015 89.7% of fuel poor occupants were in properties deemed as not being energy efficient (GOV, 2017), directly increasing the likelihood of fuel poverty.

A clear gap in the market exists to better integrate information about buildings in use into building design and management.  Better connected data sources would include building information models which contain 3D information viewed on screen in addition to data linked to the model such as room properties, building infrastructure and systems, and spatial relationships.  The model could be linked to data from various sources such as sensors and asset management systems, which would create a centralised complete picture.  Techniques could then be applied to provide meaningful advice to various stakeholders assisting in decision making processes, including a way to establish a feedback loop for continual improvement.  The feedback loop could include advice to improve wellbeing through better environmental condition insight, combined with occupancy information, or by assisting in narrowing the gap in performance by feeding back reasons for disparities.




Leading computer scientists at BIM Academy and Northumbria University, in partnership with the National Energy Foundation (home of the Assured Performance Process) and Your Homes Newcastle are collaborating on an Innovate UK funded project testing the feasibility of bringing data together into a central model that not only gives context, but allows for the generation of meaningful advice.  This feasibility study will focus on the housing association sector.  This sector was chosen due fuel poverty vulnerability, a need to be smarter with budgets driven by cuts, and improving upon their experiences, wellbeing and tenant satisfaction.

Feasibility will be tested by combining context data from a building information model with in use performance from real time environmental sensing and tenant satisfaction surveys.  Context data will include geometry information, construction materials and systems used, volume information including temperature, humidity, window open / close position, lighting and CO2 levels with additional touch sensors installed to collect tenant satisfaction.  A web portal will be developed so tenants can view a 3D model of their flat and explore current sensor readings, readings over a period and recommendations such as heating advice to reduce energy consumption or to opening a window to reduce CO2 levels to improve wellbeing.




The study will focus on four of the seven wellbeing areas which are air quality, lighting levels, comfort and mind – these are measurable through sensors and surveys with minimal intrusion to the tenants.  Wellbeing areas of nourishment, water intake, and fitness are beyond the scope of the study.  Machine learning will automatically learn and improve recommendations given to tenants and building managers to optimise energy efficiency and wellbeing.

Data collected within each flat will also be combined with facility level information such as maintenance work history, and occupancy information such as number of tenants and demographic.  This data will be combined to display a 3D model for a facility manager to view.  The model will visually highlight key information such as hot, cold or humid spots in addition to building areas with poor satisfaction, which will be linked to properties of the design.  Recommendations will be made around highlighting preventative maintenance work to reduce maintenance costs and in turn minimise the impact.




The project aim is to provide building owners and occupiers with actionable advice to optimise the performance of their building, maximising energy efficiency and occupant wellbeing.

In doing so the project will ultimately drive industry change and benefit all building stakeholders by:

  • informing owners and occupiers of their building performance
  • advising building managers and residents on how to optimise performance while a building is in use
  • enabling stakeholders to action lessons learnt throughout the procurement, design, build and handover process