Empowering Field Service Teams with Agentic AI

This report synthesises the outcomes of the industry round table on AI in Field Service held at the Agentforce World Tour London on 4th December 2025.  It focuses specifically on the Energy & Utilities sector and is based on the discussions from the tables at the event; but also integrates key insights and capabilities from supporting industry data. 

The discussion revealed that while field service organisations are dedicated to digital transformation, they are collectively challenged by administrative waste, resource shortages and legacy system fragmentation all of which underscores the urgent need for AI and Agentic technologies.

1. The Mandate for AI: Current Challenges and the Cost of Inefficiency

Field service teams are currently stretched to their limits, facing limited team visibility, delayed data, inconsistent mobile service in remote locations and non-optimised scheduling. A core challenge articulated by participants was the prevalence of manual and reactive work.  This was highlighted as manual line patrols and helicopter inspections for post-event damage detection, but also the lack of autonomy to fix what was broken due to the restrictions of the schedule and process. 

Many participants also called out the frustrations of dealing with fragmented technology landscapes, often administering in excess of 1,000 different, disconnected applications, resulting in a disconnected customer and field experience.

This systemic inefficiency places paperwork ahead of people’s work rather than the ultimate goal of not just ensuring the right person arrives at the right time with the right information, but fundamentally aiming to get the job done right the first time.

Key Challenge	Quantification from Pan-Industry Sources
Administrative Burden	68% of a technician’s time is spent on administrative tasks. Mobile workers waste an estimated seven hours per week on administrative overhead. Administrative tasks consume 15% of every workday, often consisting of form-filling instead of problem-solving.
Scheduling Failure	47% of scheduled appointments do not go as planned. Scheduling conflicts rank as the #1 productivity blocker.
Workforce Strain	56% of mobile workers report burnout. 81% of tradespeople and technicians work overtime at least once a month solely due to administrative overhead.

Source: Salesforce State of Service, 6th Edition (2024) & Salesforce Voices from the Field: Mobile Worker Research 

2. Industry Initiatives and Future Intent

The World Tour participants outlined a clear shift towards proactive, AI-driven solutions, primarily focused on Detection and Monitoring and Technician Enablement.

A. Detection and Monitoring

Organisations are moving from basic visual inspections and the use of readily available tools that can be out of date (such as Google Earth) to leveraging AI for asset intelligence.  Some examples discussed were:

• Video and Drones: Plans include using drones and video for asset inspection, such as detecting wind or storm damage and automating activities within the flow of work for damage assessment and response. This included HSE drivers such as delegating high-risk inspections to drones to increase safety standards.
• Asset Intelligence: There was significant interest in: deploying infra-red technology to detect underground damage, currently a significant time drain; use of tools like robots in the field and digital twins for proactive detection in the field; and in combining real-time satellite feeds with SCADA connected (IoT) alarms and event monitors to validate problems and incidents.
• Multimodal AI: There is specific interest in applying Multimodal AI to derive deeper information from captured images, such as scraping text from substation stickers or bar codes, or analyzing non-conforming waste to assign a safety/risk rating to avoid hazards like plant fires.

B. Technician Enablement

The most prominent area of interest is leveraging AI to reduce administrative friction and enhance support:

• Voice-to-Form (V2F): Due to engineers working in conditions requiring PPE, making screen tapping difficult, V2F technology is seen as a major benefit to save time and effort by removing manual tasks. This directly addresses the pain point of up to 30 minutes of admin and 12 forms required per job (stated by one of the attendees).
• Knowledge Transfer: AI is viewed as a solution to help bridge the knowledge gap of the aging workforce and aid younger workers in upskilling by providing visual assistance and contextual help based on image processing.
• Proactive Communication: Organisations plan to provide timely and accurate updates to customers from proactively notifications e.g. “We are aware of the leak happening in your area and are working to fix it” through to notifications covering service restoration once fixes are scheduled and completed.

3. Barriers and Constraints to Adoption

The industry must navigate significant resistance to change and regulatory complexity. Some of the key challenges from the round table discussion were:

Category	Specific Challenge or Barrier
Internal Barriers (Trust & Control)	Fear and lack of cybersecurity knowledge regarding AI, compounded by a fear of lack of control when data leaves the trust boundary. Organisations prefer descriptive rule-based systems over dynamic/adaptive systems due to the “fear of the unknown,” demanding proof that guard rails are effective.
Internal Barriers (Operational)	Dealing with many incompatible systems and multiple asset management systems. Fragmentation results in siloed data.
Internal Barriers (People)	Lack of trust in new technology due to low connections and upload errors, along with basic behavioural resistance (e.g., crew members completing forms for others).
External Constraints	Drone regulation (when and how they can be flown). Operating within a regulated industry involves major challenges.
External Constraints (Data/Risk)	Risk of large fines (e.g. up to £10k per day for Section 74 of the New Roads and Street Works Act 1991 (NRSWA)) demands perfect adherence to compliance. Third-party data, such as mapping, is often out of date.

4. Market Requirements for Next-Generation AI-Enabled Field Service Platforms

To effectively modernise field service operations and capitalise on digital transformation opportunities, organisations require robust platforms that address the core challenge of data fragmentation and administrative burden. 

The following capabilities are essential for any solution seeking to deliver scaled, intelligent field-enabled operations.

A. A Unified, Data-Driven Architecture

Modern field service management (FSM) solutions must be built upon a holistic architecture that eliminates operational and data silos, enabling a seamless flow of information between the field, asset, the back office and feedback to the customer. Given the prevalence of fragmented technology stacks, where enterprises often rely on many different, disconnected applications, a unified view is critical for transforming reactive support into proactive service.

Key architectural requirements include:

• Integration Capabilities: The platform must feature robust integration capabilities, such as open APIs and prebuilt connectors, necessary for achieving seamless interoperability with core enterprise systems like ERP, CRM, and IoT platforms. This orchestration minimizes disruption and accelerates time to value during digital transformation initiatives.
• Data Activation and Proactive Service: Solutions must incorporate mechanisms to unlock and activate real-time operational data, particularly leveraging asset telemetry and IoT devices. This capability is crucial for implementing predictive analytics, which enables the critical shift from reactive support to predictive maintenance, preventing equipment failures and optimising maintenance schedules before failures occur.
• Geospatial Intelligence: The system requires integrating geospatial intelligence with core service data to support environments managing linear assets (e.g., in utilities and infrastructure). This functionality must enable integrated geolocalised data access (GIS-enabled) and operational mapping, ensuring technicians and dispatchers have immediate, single-tap access to asset and job data in a map view.

B. Agentic AI Capabilities: Scaling and Empowering the Workforce

AI agents represent the necessary breakthrough needed in field service, transforming labour-intensive processes by embedding intelligence directly into the workflow. The goal of adopting AI must be to augment field teams, allowing human expertise to multiply without replacing technicians, thereby improving efficiency and focusing efforts on value delivery rather than administrative overhead.

Essential AI capabilities required to remove administrative and operational friction include:

• Administrative Automation: AI must handle routine tasks and streamline communication to cut administrative friction. This ranges from pre-work admin e.g. ensuring permits are in place and engineers are briefed on equipment needed for first time fix through to using technologies like Voice-to-Form functionality, which assists field engineers by converting speech into completed work summaries and forms, removing manual effort and improving technician satisfaction.
• Intelligent Scheduling and Optimization: The solution must deploy autonomous scheduling capabilities to manage high volumes of appointments, capable of optimising schedules and handling last-minute changes dynamically. Such intelligent applications are crucial for driving advancements in scheduling optimization and helping address scheduling conflicts e.g. permits, traffic management etc., which are cited as the #1 productivity blocker.
• Knowledge and Upskilling: AI should provide support that helps address the challenge of an aging workforce by providing knowledge assistance and visual troubleshooting tools to help younger workers quickly upskill.
• Safety and Assessment: Utilising AI, such as visual AI or Multimodal AI, is necessary to derive deeper information from captured images for compliance and safety purposes, such as analysing videos for non-conforming waste for safety ratings or scraping text data from asset tags.

5. Recommendations for Driving AI Adoption

Success requires a comprehensive strategy that prioritises the field worker and establishes a modern data foundation.

Recommendation Focus	Key Action Points
Prioritise Worker-Centric AI	Implement AI to automate administrative tasks (like form filling via V2F) first, addressing the workers’ biggest pain points. AI solutions should augment field teams to enable them to focus on delivering value.
Establish Trust and Training	Be prepared to move at a pace the business can adjust to, recognising that market inertia and risk aversion exist. Commit to training the workforce on AI skills.
Modernise Foundation	Overcome fragmented technology stacks by choosing vendors with robust integration capabilities (open APIs/prebuilt connectors) to ensure interoperability with core enterprise systems.
Measure Quantifiable Value	Measure success using relevant KPIs beyond basic functionality, such as Reduction in weekly admin hours per technician, Improvement in appointment success rate, and Decrease in monthly overtime frequency.

6. Proven Business Value: Quantified Success

Organisations deploying Agentforce for Field Service have achieved significant and measurable improvements, validating the investment in AI.

Success Metric	Quantified Value
Efficiency/Productivity	45% increase in efficiency.  Technicians at a global engineering firm have a comprehensive, AI-generated summary about the work required, the customer, and work order history, saving significant time per job and increasing customer satisfaction.
Cost Reduction/Opex	26% fewer truck rolls per month A US water Company deployed Asset Service Prediction and Data Cloud in combination with self-manufactured IoT devices to see their customers’ water data in real time resulting in fewer truck rolls.
First-Time Fix/Resolution	30% of service events scheduled via self-serviceProvides faster assistance by automating appointment scheduling to ensure the right service at the right time and by the right resource (roadside assistance organisation)
Customer Experience (CX)	30% increase in customer satisfaction (CSAT)70% of appointments at Global Services firm are now auto-scheduled delivering an “Uber-like” experience to customers, achieving a 30% increase in CSAT

Source: Salesforce Customer References, additional data can be provided on request

The financial drain resulting from administrative waste and scheduling failures represents a competitive advantage waiting to be activated by organisations that strategically deploy AI solutions like Agentforce

Salesforce Agentforce: The Platform for AI-Enabled Field Operations

Salesforce is positioned as a Leader in key industry assessments, including the IDC MarketScape for Worldwide AI-Enabled Field Service Management Applications and Worldwide Field Service Management Solutions for Utilities. Furthermore, independent assessments categorise Salesforce as an Exemplary Provider and a Leader across critical product and customer experience metrics in Field Service Management, Customer Engagement, Proactive Maintenance, and industry-specific FSM Buyers Guides (Manufacturing, Power & Utilities).

The cornerstone of this offering is Agentforce for Field Service, which transforms operations by embedding predictive, generative, and agentic AI directly into the flow of work.