Insight

AI opportunities for field services

Luke Matcham

Any organisation that employs a field force will be aware of the myriad challenges around organising employee routes for the week, ensuring they have the right tools and equipment to complete their jobs, and dealing with last-minute changes to task priorities. For years, software providers have claimed their solutions can deal with these challenges and offer significant benefits. Still, recent advances in artificial intelligence (AI) technology are starting to unlock new opportunities and make existing benefits much more tangible.

How do you separate the hype from the reality and determine which AI tools and solutions are appropriate for your organisation?

AI implementation

Our recommended approach

Step 1

Define the AI vision and guiding principles

Step 2

Identify AI use cases and define a strategic blueprint

Step 3

Prioritise pilots and define an AI roadmap

Identify potential AI use cases within your field force function

Field force functions can vary drastically from one organisation to another, e.g. one organisation deploys engineers to fix broken-down assets while another has a fleet of travelling salespeople. While this article focuses more on asset-related field force activities than those associated with customer service/sales, there are many common characteristics and challenges that AI use cases look set to tackle.

Promising recent AI developments for field services

Workforce planning optimisation

Typically, organisations suffer from significant periods of field force operative downtime caused by inefficient route planning.

Machine learning can optimise route planning to deliver significant efficiency benefits by incorporating feedback from various sources, such as typical traffic and weather patterns. Real-time decision-making capabilities from AI can also be used to re-route operatives to the highest importance jobs when they occur, without human intervention.

Remote diagnostics

If a field operative encounters an issue in the field which they were not prepared to deal with, e.g. they cannot find the job location or they are unable to diagnose what is wrong with a broken-down asset, this would traditionally result in the need to call back to base for further information.

AI diagnostic capabilities and chatbot functionality will assist with this. For example, it could provide step-by-step instructions on how to carry out a repair or automatically analyse historic job/asset data to determine a likely cause of failure.

Job debrief automation

Collating accurate debrief data can be a perennial challenge for organisations and a significant draw on operatives’ time spent in the field. In many cases it may even be unsafe to stop and record notes whilst working on machinery so the option of transcribing via an AI voice recognition tool can offer significant benefit.

Job records can also be automatically augmented with location and time data captured from an operative’s device to provide additional information about the likely time required to complete future tasks. If these records are stored in a knowledge management system, AI can share relevant content on future jobs and ensure that employees learn from each other’s experiences.

Optimising inventory management

Generative AI can forecast inventory needs accurately in advance, optimise warehouse stock levels, and lead to more efficient inventory management. Predictive analytics can also suggest which tools or equipment may be required to complete a job and build stop-offs at relevant depots into an operative’s daily schedule to give the best chance of success upon arrival at the job.

Predictive maintenance

AI can effectively use the vast data sets collected from jobs by predicting when assets will likely fail or encounter issues. This can generate predictive maintenance plans that actively maintain assets before they go wrong, using data such as hours run and historical trends on asset classes or even specific assets.

Evaluating feasibility and implementation risks

Dedicated field service management software incorporates increasing AI functionality, yet several bespoke tools also exist that may tackle one or more of your organisation’s challenges. A good evaluation of available AI offerings will encompass both categories and weigh the potential benefits against both organisational and technical feasibility challenges.

Organisational feasibility factors include:

User adoption, which may be challenging for longstanding field force members who have set ways of working
Developing the internal technical skills required to run your AI initiative after implementation
Strategic alignment to wider business goals rather than tactical deployments
Potential need for a new operating model to take advantage of AI benefits
Pushback from users who may feel threatened by a collection of location data.

Some of the technical feasibility challenges include:

Lack of human decision-making capabilities, e.g. a planning system could prioritise based on job type criticality but would not be able to replicate the operator’s local knowledge and experience
Lack of supporting technology, e.g. advanced telematics devices in vans or accurate GPS tracking on all user devices
Strong and stable internet connections, i.e. all field-based solutions are likely to require a solid connection to be effective, but internet connectivity remains limited in large parts of the world
Ease of access to, and quality of, the data required, i.e. organisations may be guarded over their data and hesitant to share with others
Complexity involved in tailoring generic AI solutions to your individual business needs.

Leaders implementing AI initiatives should also be aware of external factors, such as rapidly changing regulations around AI, which may impact the direction of any initiatives. AI use within a field force setting is also open to potential challenges from operatives or unions who may question the need for and use of personal data about employee whereabouts throughout the day, for example.

Calculating the return on AI investment

Determining the best use cases for AI within your organisation can feel overwhelming. Developing a business case for any AI initiative, which weighs potential benefits against the risks and costs, can help you make an informed choice by comparing which are likely to return the most value from your initial investment.

Key benefit categories for an AI business case

Improved employee safety

Of utmost importance for any organisation, any initiative which can improve your employees’ chances of arriving home safely at the end of the day should always be considered. AI advances in this area include:

the ability to monitor adherence to safety checklists in real-time
the ability to provide operatives with automatic updates about potential hazards from image recognition
knowledge management based on historical jobs
data around severe weather warnings.

Improved operational efficiency

Operational field force efficiency can be improved by minimising employee downtime between jobs due to:

improved routing
automating the dispatch of jobs in real-time
improving first time fix rates by predicting the skills, tools and equipment needed to complete a job
significantly speeding up the job debrief process.

The array of potential efficiency improvements is vast and should form a substantial part of your case for change.

Sustainability

Improved vehicle routing means fewer carbon emissions from unnecessary travel between jobs. Improving the first-time fix rate also results in reducing the number of repeat visits required, resulting in less travel. If your initiative results in the collection of new data from the field, this may also make it easier for your organisation to report on carbon emissions and determine where to apply other initiatives to reduce its carbon footprint in future.

Improved customer satisfaction

For organisations serving an end customer with their field force, such as a utility provider or repair company, improving the first-time fix rate with AI results in improved customer outcomes. AI can also be applied to determine customer sentiment when an issue is raised or use historic data to highlight and prioritise customers with repeat issues who may be more likely to raise a complaint.

A pragmatic approach for deploying AI within your field force

Any organisation aiming to roll out AI within its field force needs to take a pragmatic approach to its implementation.

This means that trialling AI initiatives on a smaller, or regional, scale may be a safer initial approach. Such an approach allows the benefits and risks to be tested without subjecting the organisation to the undue risk of negatively impacting their entire operational force. Trialling may also be a pragmatic option given the speed at which technology advances in this area are being made. Embarking on a multi-year programme could mean that, by the time it has been implemented, the technology and related benefits are already behind the latest developments.

Always ensure that your field force is bought into the change journey for any AI initiative, as end-user failure to adopt the solutions could spell disaster for any forecast benefits.

Ultimately, you should treat the implementation of AI within your field force as you would any other project. If you align AI use cases with your overall strategy, balance the potential risks and rewards, and don’t neglect the importance of effective change management, then AI looks set to radically alter the way that the field force of the future operates.