Analyst Perspectives

Field service is an often-overlooked area of the service equation, in part because managing its operations involves tools and activities outside of normal contact center models. Field service has a communications component, which is partly handled by contact centers. However, the bulk of field service management activity has to do with organizing the work of technicians and dispatchers. It also requires constant attention to informational resources about products, assets, locations, real-time conditions and, of course, customers.

Use cases for artificial intelligence fall roughly into two categories. One is the group of processes that organize field service from behind the scenes: optimizing the workforce, for example, in terms of shifts or travel, or making sure that a truck roll contains equipment relevant to the specific cases on the manifest for that trip. The second category generally tackles the execution of service contacts: What knowledge resources does the technician draw on to solve the problem, and how do you minimize the time needed to complete the job?

AI is impacting this complex mix, as it is everywhere. Some of the best use cases for field service AI are similar to those created for contact centers, and some are unique to the field service environment. In both environments, the goal is to use automated technology to ensure that customer satisfaction is high.

For example, services begin with interactions—either communicative ones or those that occur on-site. All interactions require significant amounts of process control and information and often begin with a customer connecting to a chatbot or automated agent to share information about a problem or need. The next steps are to assess the situation, determine recommended actions and urgency, create service tickets and confirm customer entitlements.

AI supports these tasks, but these functions just scratch the surface of what is now possible and emerging from field service management applications. In nearly all service cases, customers’ problems need to be analyzed to diagnose the issue, identify potential solutions and compare the issue to past cases across the customer base. These are the things that automation does well, especially scouring knowledge resources to find patterns in problems.

Increasingly, AI is applied to the more complex problems of orchestrating the details of interactions. It can predict how long actions will take and help position technicians and their parts, tools and equipment for optimal response times. Field service processes have more moving parts (literally) than contact center processes, and AI is now being tuned to optimize the movement of pieces on the board: grouping work orders by location or priority, for example. All functions that depend on improving efficiency through awareness of resources and context are strong candidates for both automation and AI intervention. By 2026, 4 in 10 enterprises offering field service will start AI automation pilot projects to reduce dispatch requirements and proactively engage customers early in the service process.

There are also benefits to applying AI to the work of the technician onsite. The field service industry has talked about augmented reality for some time. Prospects for wider adoption are strong now that software providers are using AI and machine learning to visualize complex objects in two or three dimensions. Applications for AR technology go well beyond field service, including manufacturing, automotive, medical and energy industry applications that are spurring development at a rapid pace.

AI also maximizes onsite diagnostics, particularly in combination with predictive analysis of products in situ based on sensor data or Internet of Things information. When used in conjunction with AR and mobile applications for information retrieval, onsite technicians can connect to knowledge bases at the home office, as well as work order details and remote assistance (from humans or automated sources). This makes it faster and more accurate to understand problems and work towards solutions, with better communication to the customer along the way.

There are also behind-the-scenes processes that impact the quality and speed of field service interactions. As noted, AI can boost the efficiency and effectiveness of pulling solutions out of knowledge resources. That’s a pretty standard starting point for service delivery in any context. Field service organizations, like contact centers, are also ripe for process improvement in labor management. The dispatch process has a lot in common with the workforce management needs of a contact center—for example, forecasting interaction volume, mapping resources to needs and adjusting based on novel contingencies.

In the case of field service, AI can improve how schedules are produced and updated, along with optimizing resources like fleets, parts and tools: the entire process of deciding who should be where, when and with what supplies and information. AI is also being used to dynamically adjust scheduling decisions based on criteria like traffic, weather or new urgent requests. It is especially good at route optimization in real time.

Many of those applications fall into the category of improving processes that already exist (and are already pretty efficient). More exotic uses for AI are emerging that might create entirely new paradigms for how complex service requests are delivered. One example is using automation to analyze and understand the context around failure modes in complex products. AI models can be trained to understand where and when failure points emerge and apply that information to other assets where solid data on failure rates may not exist. AI models also have a role to play in assessing sensor data from relatively small training samples to understand how to ensure the continuous functionality of expensive equipment that is hard to service. This can speed the time to value for asset health prediction.

AI is not a magic bullet for field service delivery. It doesn’t turn bad processes magically into good ones or create resources where none exist. However, it does provide ways to maximize the outcomes of the resources that do exist and to create new, more efficient processes. As in other areas, service managers are still discovering which processes lend themselves to AI improvements.

Many enterprises would like to use their field service operations to generate revenue. This is hard to do when under constant pressure to control costs and deliver top-notch outcomes to high-value customers. Using today’s field service tools that enhance automated processes with AI is a way to turn these risky and complex service interactions into brand-building encounters. When a business can predict a part or machine failure and automatically alert the customer (or fix it remotely), it builds customer confidence and helps justify service relationships and high-end service contracts.

The use cases businesses should explore depend on many factors, like the complexity of what’s being serviced and the existing technical maturity of the service organization itself. In other words, an organization with a mature service environment (especially around its contact center) may use tools that lend themselves to AI intervention, like agent guidance, knowledge management and remote diagnostics. These naturally lead to using AI for tasks specific to field service, like automating dispatch and helping customers articulate the true nature of their service needs.

Enterprises should be aware that the technology landscape around field service is changing very rapidly. Take stock of the tools used in existing processes and their effectiveness. In evaluating the current generation of field service management tools, enterprises should consider how software providers use AI, for which use cases and how well those providers articulate the ROI and benefits of the technical advances.

Regards,

Keith Dawson