Applying AI and Chatbots to council planning applications

Applying AI and Chatbots to council planning applications 1200 675 Agile Datum

There’s a wealth of information available to us online 24 x 7 but are we better informed?  Possibly not when it comes to specialist areas such as planning law.  Unless you’re a serial house owner/mover, live on the doorstep of a new development or are troubled by a neighbour’s intended plans,  the chances are that you’ll come into contact with the planning legal system pretty infrequently.

According to the BBC, people used to move 4 times after their first purchase, but this has reduced to just 2 times recently since the introduction of simpler permitted development rights, and owners taking the opportunity to alter or extend their homes rather than move.

According to the Ministry of Housing, Communities and Local Government statistics, permitted development applications increased by 30%.

There’s a wealth of information on local authority websites and on about the many different types of domestic permitted development , so theoretically, it should be easy for citizens to “self-serve”.  However, In practice, the rules over dimensions and materials, concern that plans may affect the setting of a listed building for example or the risk that the local authority may have withdrawn specific rights mean that owners will want to reassure themselves that permission isn’t needed. This has increased planning officers’ workloads, and in fact, more than 50% of applications were not needed.

Local authorities are doing their utmost to make planning content straightforward and easy to locate on their websites, however, citizens often complain that the call centre can’t help, they can’t get through to an officer or that their emails go unanswered.

Could Conversational AI (Chatbots) be the answer?

Whilst it’s generally true that planning is so varied that no two cases are the same, householder permitted development has very clear processes and programmatic rules that make this area of planning a prime candidate for AI.

Self-learning chatbots offer authorities the potential to improve the user experience from day one and are increasingly effective over time.   Conversational AI, well-implemented, could provide instant responses to an estimated 40%-50% of “standard” planning related queries across multiple channels. To ensure that the user experience is positive, conversational AI can provide proactive updates and notifications. Bots can also be trained to hand over to a human when an interaction fails or the citizen’s responses suggest that this is not the ideal route for their particular issue.

It’s been estimated that in one unitary authority serving a population that’s projected to grow to 300,000 by 2025, combining chatbots with other machine learning processes could save over 5,000 man hours yearly (the equivalent to having 3 extra “virtual” planning staff per council).   That’s time that could be focussed on an ambitious home building target that may be difficult to reach otherwise given current planning staffing levels.

A strategic choice for a great customer experience

Today there is a vast quantity of data across organisations.   At the same time, the cost of cloud computing power has fallen significantly.  The availability of lower cost cloud infrastructure combined with  the emergence of new AI tools and algorithms that can handle increasingly complex tasks all contribute to your organisation’s ability to provide a better customer experience 24/7/365. The technology is available – and affordable.

The world’s largest cloud software/service providers have invested significantly in conversational AI seeing this as strategic to their customers’ digital service delivery.  Some are developing hybrid systems which combine handling  “short tail” enquiries (high certainty responses to a repetitive and limited set of service users’ questions/tasks) with “long-tail” behaviour (providing multiple possible answers generated automatically from supporting documentation for infrequent tasks).

Planning is very document “heavy”.  Microsoft visual image recognition enables images to be processed as data and incorporated in chatbots, adding another dimension to the user experience.

The level of planning related enquiries coming in to a council will depend on many factors.  As well as those stemming from major applications for housing, windfarms or industrial development,  householder applications for small scale development can generate high call volumes from neighbours. A typical council can receive around 8900 planning specific call centre enquiries per year (doubling to some 18,000 queries per year when you include email and web forms).  It’s estimated that 40% of these interactions could be closed on the first call using the power of chatbots.  That reduces the number of interactions by around 7000, freeing at least 300-400 man days of capacity.  As the chatbot is trained to handle more complex interactions, this additional capacity can potentially double. More importantly, being able to interact with a chatbot 24/7/365 gives the citizen a more immediately productive experience.

UK Local Government planning authorities currently receive over 3,500,000 applications every year.  At the same time, there’s been a 14.6% fall in real-terms in expenditure on the planning function and a 15% overall fall in staff numbers between 2006 and 2016.  Some 117 of 400 councils have missed their housebuilding targets by upwards of 65% over the last 3 years, and nationally we are only achieving 300,000 new builds versus a government target of 450,000.

If data is locked away in non-machine readable, unstructured documents filed in proprietary file management systems, planning officers will find it hard to access any form of aggregated local data for strategic insight.  Combining a fully digital planning solution with Microsoft machine learning from Agile Datum, for process automation and AI tools, with the Agile Datum chatbot at the forefront of customer service, will free up planners to focus on house building and commercial developments that are very much needed.

Where do we go from here?

Across all industries, it’s believed that 80% of an average working week is focused on low value, high volume admin and only 20% of the week focused on high value, strategic, customer interactions that support growth and positive outcomes.  Agile Datum aim to maximise the use of AI, machine learning and bots to streamline background administration tasks and support skilled employees to reduce mundane admin, freeing them to spend upwards of 80% of their time on high value activities and citizen engagement.

An Agile Datum rapid proof-of-concept project can demonstrate how these technologies will work with your data and your organisation objectives.  We’ll help you to design an intelligent solution to improve business outcomes for a high-value business problem.

Contact us via our website or learn more about the Rapid Proof of Concept packages on the Government Digital Marketplace.

%d bloggers like this: