Source – information-age.com
While Machine learning has been on the technology agenda now for as long as twenty years, it is only in more recent times that its potential benefits in field service management terms have been more widely understood.
Machine learning is an offset of artificial intelligence (AI), whereby AI can learn to do a task without needing to be programmed by a human—the AI learns to do the task itself. There has been growing interest in AI in recent years, with technology leaders, including Elon Musk and Mark Zuckerberg, utilising machine learning to enhance existing technology.
The excitement for machine learning in service organisations has increased profusely, even though its powerful benefits are not completely understood by the masses yet.
Organisations around the world are now beginning to see machine learning as a ‘predictions-as-a-service’, whereby all manner of macro and micro environmental data—such as weather patterns and a specific technician’s capabilities — can be seamlessly connected and analysed to provide accurate predictions based on history.
And, while all this can be done without any interpretation effort required from professionals working in service-based organisations, the actionable insight it delivers can create significant competitive advantage.
But is the machine rendering man redundant?
Quite the opposite. In fact, the combination of machine learning predictions and operational research carried out by business leaders is providing a deeper and highly valuable level of business intelligence that is enabling more informed strategic decision-making, as well as improved productivity and performance.
For example, with machine learning technicians can become increasingly efficient in a way that makes businesses profitable, while also reducing the time that is spent on some tasks.
So how exactly is machine learning leveraging new opportunities for field service organisations?
When it comes to delivering business value through machine learning, the primary opportunities revolve around better planning and more precise scheduling. Three key examples include:
1. Predicting traffic patterns
Innovative service organisations are now introducing the ability to route engineers and technicians according to predictive traffic patterns. Based on historical data such as bank holiday traffic patterns, this enables technicians to be dispatched to particular jobs when traffic is less congested in those specific locations.
This is already delivering huge time and cost savings, not to mention improvements to the customer experience by reducing delayed arrival times and the need for long ‘wait-in windows’.
Beyond this, ML derived predictions can also provide organisations that deliver repair and installation services to the home with more accurate indications around the estimated duration of specific jobs, enabling scheduling and productivity to be further optimised.
2. Weather forecasting
The UK Met Office holds climate records dating back to 1959, so with 60 years’ of historical weather data, the Met Office has developed a weather forecasting model that enables it to predict weather patterns based on historical information and other seasonal factors.
Similarly, field service organisations are starting to mirror this weather predictions model by adding machine learning capabilities onto their existing management systems.
This is delivering quick time to value by identifying when certain jobs—often those that need to be performed outdoors or at height— should be postponed due to the expectation for poor weather and associated health and safety concerns, as well as time and cost considerations.
3. Preventing customer no-shows
One of the greatest profit drains for businesses operating in a field-based context are customer no-shows whereby a technician travels to a customer’s home at an agreed appointment time only to find there is no one at the property to provide access. Unsurprisingly, this is a key frustration for businesses.
So how can machine learning help with this? It can better predict whether a specific customer is going to be at home or not based on historical data about the specific customer’s track record, the location of their home and a host of other factors relating to the weather and their work situation.
The potential for this actionable insight to eliminate wasted technician time is significant and is expected to provide an increasing source of competitive advantage.
4 ) Sending the right person to the right job
Machine learning can also streamline service offerings by allocating certain engineers to specific jobs. For example, if an engineer frequently installs smart meters into homes, that engineer will become familiar to that job and will inevitably become faster and completing the installations.
Because of this, machine learning software can reallocate that employee for future smart meter installations to speed up job processes. Streamlining business decisions through machine learning can ensure that employees can work on the jobs that they excel at, improving customer satisfaction.
5 ) Predictive maintenance
By leveraging the data that is generated by the Internet of Things, machine learning can anticipate when repairs will be needed and can proactively schedule service without requiring human intervention.
Consequently, machine learning can monitor the status of the equipment and can predict when an issue may arise, allowing engineers to attend to the equipment before the issue is encountered.
By utilising preventive service over reactive service, organisations can prevent costly failures and can stop spontaneous breakdowns that irritate customers and take up engineers’ time.
How the machine is driving customer and employee experiences
From the viewpoint of the customer who needs their boiler or dishwasher repairing, the benefits of machine learning can include a dramatic increase in ‘first-time-fixes’ by ensuring the right parts and technicians are dispatched to them first time.
This is improving overall customer satisfaction and experience levels— something that is becoming increasingly critical in an environment where customers now demand Uber-like service levels and have more choice and sway than ever before.
Similarly, for employees working in field service organisations, machine learning can also improve the overall employee experience and support staff retention levels.
Dispatchers’ day-to-day roles are made easier by the fact that they have fewer decisions to make on their own, and for engineers and technicians, wasted trips can be minimised by reducing instances of customer no-shows or arriving at a job without the requisite parts or skills.
What’s next?
Of course, machine learning remains a new concept for many and questions remain around how best to apply this in a field service context. There is also still work to be done to tie machine learning into existing workflow systems so that future predictions can be more easily integrated, understood and applied.
The organisations that master this ahead of the masses, however, will certainly be able to improve their compliance with Service Level Agreements (SLA) and reap better ongoing business rewards.