How can AI help development assessment?
We recently had the opportunity to use OpenAI's latest version of it natural language processing tool Chat GPT4. One of the interesting developments in this version is the ability to upload images and ask the engine questions about the image which is a huge step for developing meaningful business cases for its use.
Now its important to acknowledge that there are many ethical questions and dilemmas that need to be tackled about the broad use of AI, even the development of ChatGPT has brought to surface allegations of mental health impacts on those staff who train the engine by watching, analysing and manually processing images, many containing abusive and confronting material. There is some sobering reporting on this in the Al Jazeera Podcast The Take.
However it's also important to highlight some practical examples of where AI is headed in useful and practical terms. In construction environments, the proposed building now existing more often than not in a digital form. The typical neighbourhood house right through to large commercial buildings and sky scrapers will be drawn in a digital model instead of paper. This model is not only quicker and cheaper but is spatially and locationally correct, meaning that the movement of the sun and its heat load on the building can also be accurately modeled.
The benefit? If the client wants to add a room or move a window you no longer have to redraw the plans, just drop and drag a new part from the library of models and bang it's good to go. But the digital workflow stops at the sign off of final plans. Once it is sent to the planning authority its either a PDF or good old paper for the assessment.
Whilst the planning assessment and construction permitting process will eventually become a native digital process and authorities will work out how to handle digital models, there's a good chance AI will bridge the step in-between.
The words 'properly made application' will make any assessment officer sigh, Councils across the country spend huge amounts of time pre sorting and determining if applications have been put together correctly. I have seen plenty of clients plans sent back because the North Arrow disappeared when they were printed. So if properly made applications are such resource drain, we wondered what would happen if you uploaded a plan into Chat GPT4 and asked it some questions.
Well AI isn't going to replace the Council assessment officer anytime soon, but we were surprised by how ChatGPT4 once prompted could recognise building setbacks, listed dimensions and even the north arrow. Even ChatGPT3 which is available for public testing has managed to digest and analyse a large number of Plannings systems allowing you to ask basic questions. However a general language AI isn't going to rock the world of development assessment just yet, that is more likely to come in the form of a specially trained AI which understands the regulatory framework like the well known robot lawyer 'DoNotPay' and some carefully crafted regulations to help launch it.
For more information about how AI and other technology can help your work, sign up for our newsletter: