Martyn Stones outlines details of our latest pilot project in association with Oxford Brookes University, to aid damp, mould and condensation detection.

I suspect few would argue with the view that current stock condition assessment methods are time consuming and labour intensive. With increasing regulatory requirements, not least the implications of the forthcoming Consumer Standards regarding safety and quality, combined with constrained working capital, surely a collaborative approach to make better use of data and technology could provide widespread efficiency benefits for the sector and, more particularly, support the wellbeing of residents.

Since the start of the year, Oxford Brookes University (OBU) has been working closely with Andrew Burke and other members of the NHMF Management Team on how best use could be made of its capabilities particularly regarding artificial intelligence, machine learning and computer vision. The best research addresses real world challenges and does not simply sit on a shelf somewhere!

Consequently, we agreed to carry out a discovery exercise across the sector to understand the appetite particularly for the use of computer vision, where OBU has a particular specialism, along with any other gaps in requirements where we might be able to add value.

In short, there is a lot more going than originally anticipated. It is certainly evident that there is an appetite for AI based computer vision to assist with defect diagnosis, repairs, building safety and asset management but also potentially linked to predictive modelling as well. Alongside this, standardisation of data in terms of defect categories and their scoring will also be required to allow like for like comparisons and linkage with other key datasets such as the Schedule of Rates. As already suggested, if we are to make the progress required there is a need for collaboration and a more strategic approach, not least to eliminate duplication of effort as much as possible. We have seen that here is a clear willingness to work together and perhaps not surprisingly, interest also extends to the private rented sector as well. There is also the potential benefit to the Exchequer, both in terms of healthcare implications and potentially reduced regulatory costs. We are therefore hopeful that our work may also attract the support of Central Government, the Regulator, and the Ombudsman as well.

We are now working with one of NHMF’s largest landlords to pilot computer vision with a view to automating the detection and categorisation of dampness, mould and condensation in digital images. To achieve this, we shall be working with one of the landlord’s surveyors to help train the algorithm that has already been developed so that it can mimic the professional judgements made by the in-house team. To start, we are considering three potential ‘use cases’ to assist with the following:

1. Reporting and triage of repair requests via images taken by residents’ smartphones and linked to the landlord's online customer portal;
2. Allocation of tradespeople along with equipment and materials to improve first time fix rates and reduce the number of return visits; and
3. Quality control of work completed and comparison against original claims.

We shall keep you up to date on progress in future blogs and in between times, if you are interested in learning more about our work, please feel free to contact me by email: mstones@brookes.ac.uk