More and more housing providers are turning to enhanced AI to strengthen their operational performance, improve customer service and retain skilled staff.
Chatbots began to gain popularity in the mid-2010s. The technology and its applications have significantly improved, making it more accessible and useful now that it’s moved from simple forms of automation to more complex queries.
Chatbots are an example of AI working for both tenants and housing providers simultaneously; today’s AI solutions are addressing a specific rebalancing, from ‘pure play’ customer service improvements to the transformation of operational performance.
For scheduling specialists like us, AI has the power to transform field-service management through our proprietary algorithm and machine-learning capability. The underlying ideas behind it are ‘how can housing providers achieve the optimal field-service plan?’ and ‘how can they solve scheduling and in-day control of routing in the most resource-efficient and SLA/KPI-oriented way, and without having to wait for the answer?’.
Millions of solutions
This is a highly complex mathematical and logistical problem. If you analyse the optimal sequence of 10 field-service appointments for a single operative, there are 3.6 million solutions. Other factors to consider include time constraints, fixed appointments and breaks, order specifications (such as skills, tools and stock) and human factors (such as illness and cancellations).
FLS’s solution, the PowerOpt algorithm, was developed and evolved over 25 years of specialist focus and has matured as the market has transformed. The algorithm is the core intelligence for optimised scheduling and dynamic route planning, taking into account all factors in seconds, scheduling appointments, employees and materials.
The PowerOpt algorithm makes it possible to control logistics and service processes in the field in a cost-optimised, sustainable and customer-oriented way, with complete transparency over the resulting outcomes. AI calculates in real time, including predictive traffic on each road segment for the time of day, through continual optimisation of planning and ongoing coordination without needing further intervention.
AI-supported field scheduling
AI-optimised field-service management offers a solution to many long-standing housing problems, including the reduction of ‘no access’ visits, planned maintenance backlogs and dealing with voids, with the resulting data supporting compliance and tenant-satisfaction reporting.
FLS’s AI enhances scheduling accuracy through sophisticated predictive analysis. This enables intelligent predictions about appointment durations and arrival times. Thanks to this machine-learning approach, appointment accuracy is continuously improved. Its advantages include:
- ‘What if’ simulations for optimum capacity and resource scheduling;
- Increased up-to-the-minute appointment accuracy using geo-coding and predictive traffic for specific journey times;
- Improved scheduling accuracy for linked appointments and follow-up visits;
- Increased customer satisfaction through features such as enabling operatives to begin shifts from home locations and built-in depot visits.
FLS is Microsoft’s scheduling partner for Dynamics 365, seamlessly integrating with Microsoft’s customer engagement and field service software. Microsoft recognises FLS Visitour as the best-of-breed field scheduling software, extending beyond the capabilities provided by Microsoft’s own resource scheduling optimisation (RSO) software. The FLS solution is also available as an upgrade with most of the leading housing management systems.
Risks and human oversight
However, there are dangers when developing AI algorithms, including compounding human bias (should it exist) in collected data and modelling, so human oversight is needed to ensure compliance with regulation and overall fairness.
Intelligent field-service automation is used to reduce response times and provide more precise information for customer enquiries; there could be data-privacy risks for tenants here, and evolving AI regulations, such as those promoting transparency and the responsible use of machine-learning models, must be adopted before any blanket bans prevent AI’s potential.
What’s next?
Predictive analytics evaluates historical data using mathematical methods to discover trends and patterns and incorporate them into calculation models for future predictions. Housing providers are already using these to identify trends to counter damp and mould, optimise energy management and help with sustainable housing developments.
The best-known methods for these evaluations include decision trees, regression and neural networks. While decision trees and regression are relatively easy to model, neural networks require much more effort. They can be represented using AI and allow very precise recognition of patterns and trends in real-time. However, they can only be used effectively if a corresponding volume of intelligent data is available; intelligent data is created with clear objectives, diverse and representative sampling, testing and regular reviews.
Innovate today
Although some technology leaders have called for a pause in developing powerful generative AI models while potential threats are explored, the power of innovation from using AI is available today and should be embraced by the UK housing sector.
FLS uses AI to generate the best field-service results to match specific circumstances, typically achieving results in a split second. With an AI-powered scheduling and routing solution, more appointments are enabled and completed, benefitting both tenants and operatives.
Lee Hawkes is the housing sector lead at FLS – Fast Lean Smart.
