A year on from our last Service Industry projections, Artificial Intelligence (AI), and now increasingly agentic AI, continue to drive transformative advances for service providers.
AI has the potential to further revolutionize service delivery through 2025 and beyond, enabling autonomous, intelligent and context-aware interactions that can provide personalized responses and go beyond that to execute tasks.
Mark Brewer, VP Service Industries at IFS, shares his insights for the Service sector, with particular focus on service providers such as those in Property and Facilities Management, Managed Services (MSP’s) and Testing, Inspection and Certification (TIC) use cases.
PREDICTION #1: BY 2027, 75% OF SERVICE PROVIDERS WILL DELIVER SERVICE AT THE SPEED OF CONVERSATION
The role of AI has continued to evolve at a breathtaking pace. We are now all familiar with copilots and chatbots, and their ability to support or enhance human productivity and effectiveness. Trained on data from specific domains, copilots use Large Language Models (LLMs) to fetch information from defined sources, requiring human input, instruction and help to execute tasks.
However, the next phase of AI introduces entirely autonomous capabilities in the form of AI ‘agents’. This so-called ‘agentic AI’ is very different; it can work independently and make reasoned decisions, and, increasingly, will be able to replace the human input needed to complete certain tasks. By using machine learning, agents improve their performance over time and can independently access a far wider range of resources.
Agentic AI is trained to make decisions by applying causal logic, like humans. Crucially, since it can understand context, it can consider complex and nuanced requests providing accurate and relevant responses.
For instance, for a customer requesting the cost of a service contract upgrade, and a quote by email, an agent would be able to check and understand a customer’s history, the specific details of existing service contracts, and the technical specifications of their equipment to automatically generate and send a quote.
Most chat bots like ChatGPT generate an answer to a query based on a Large Language Model. However, faced with a technical or specialist question, the LLM will have limited capabilities due to lack of specific data. Retrieval-Augmented Generation (RAG) allows generative AI models to be supplemented by fetching information and data from external, citable databases, knowledge bases, and web pages.
For Service industries, this means chatbots and AI agents used by customers or service technicians can access machine OEM knowledge bases and records. This creates an intelligent, trustworthy, relevant conversation, reducing the risk of AI errors and hallucination.
PREDICTION #2: BY 2027, 50% OF SERVICE PROVIDERS WILL TAP INTO A WHOLE NEW POOL OF CONTINGENT, FLEXIBLE WORKERS
Faced with a major global skills shortage, the service sector is recognizing the need to look towards greater workforce flexibility to plug the gaps.
For example, an IFS customer in the cleaning services sector is trialing an approach that sees a small team of operatives arriving on site together in a vehicle, but then dispersing to visit individual customer locations in nearby blocks and premises on foot or via cargo bikes. The model is made scalable via AI powered scheduling optimization, which can consider the makeup of crews as well as differing modes of transport.
This deployment flexibility enables a contingent, more inclusive, more sustainable workforce. It enables the hire of new part-time or flexi-time employees that may not be able, or want, to work full-time or even gig-economy workers.
Using a smartphone or tablet is second nature to most of the working age population. And these days most service workers are fundamentally digital workers. With basic task-based training, many services can be delivered very efficiently using this largely untapped human resource model.
In this way, service providers are able to leverage a contingent workforce outside of traditional domains, further driving their DEI goals, whilst at the same time addressing the issue of the aging workforce and the skills shortage – a huge win-win.
PREDICTION #3: BY 2026, FULLY AUTONOMOUS SCHEDULING WILL DEMOCRATIZE MOBILE WORKFORCE MANAGEMENT FOR ALL
Despite the advantages and efficiencies offered by Scheduling Optimization technology, many service providers currently perceive it as too complex to implement or too sophisticated for their needs. Historically, automated engines have required extensive manual fine-tuning to set and adjust the correct parameters for the system algorithms to make optimal planning decisions.
But with embedded AI, such as that used in IFS Planning & Scheduling Optimization (PSO), workforce scheduling is becoming self-learning and fully autonomous. Just as a self-driving car has been trained to navigate without human input, fully autonomous scheduling can manage workforce schedules based on dynamically adjusted value curves and real-time data, learning from past outcomes.
Low-to-no manual intervention is required to optimize rules; trained with some data, the AI learns the required decisioning behavior itself. It can become operational and value-add with minimal setup. AI is therefore making dynamic scheduling accessible to all, regardless of their size, complexity, resources or volumes, allowing every service provider to benefit from increased productivity, reduced travel and increased SLA compliance – which in the past may have eluded them.
PREDICTION #4: BY 2026, 80% OF KNOWLEDGE ARTICLES WILL WRITE THEMSELVES
Service Providers in Facilities Management, Managed Services, and Testing, Inspection and Certification (TIC) organizations need technicians with the ability to work on equipment from multiple brands spanning hundreds of different models.
Be it servicing domestic boilers, upgrading data center racks, repairing retail self-service scanners or auditing factory production line safety, service providers have relied on specialist experienced operatives certified in these processes.
The skills shortage, however, has changed this dynamic. New entrants are typically less experienced, need to become generalists, and so need greater technical support on the job and during their onboarding.
Access to an up-to-date knowledge repository is necessary, but not sufficient.
Now, with the help of AI, organizations can auto-curate knowledge articles by observing the resolution steps, spare parts and technical procedures leveraged by service technicians, augmenting this with engineering data provided by OEMs. This reduces the need for manual updates, which requires discipline and significant investment, and ensures that knowledge repositories are always up-to-date, reflective of lived experiences
Here, AI can democratize knowledge across the workforce, providing expert assistance and support with knowledge articles that write and continually refine themselves. Even better, AI translation capabilities mean the same repository can be available in multiple languages, ensuring technical information is accessible to a global workforce.
PREDICTION #5: BY 2030, EVERY ASSET WILL HAVE A FULL-TIME AGENT
Today’s devices and equipment are increasingly connected, sharing data with OEMs and service providers in real-time. With the advent of agentic AI, these data streams can be continually monitored and analyzed, allowing the agent to make observations and take pro-active actions to maintain uptime and performance autonomously.
For example, an AI agent may detect a fall in pressure in a consumer gas boiler. Based on the knowledge repository and specific model details, the agent sends an alert to the user’s smartphone, along with an explainer video link showing how to repressurize the system.
In another scenario, where a part is about to fail, an AI agent might raise an alert, but also give a customer appointment options and pricing for an engineer to attend and fit the correct with a replacement part, maintaining the boiler warranty.
For this reason, both equipment OEMs and independent service providers are investing in developing and offering their own agentic AI solutions across their customer base at pace. For service providers in particular, agents represent an additional value-add subscription revenue stream, reduced costs (fewer field service visits needed) and a way to engender ‘stickiness’ (creating customer loyalty and trust) through pro-active engagement and an improved customer experience.
In summary, AI in its various forms (Gen AI, Agentic, Time series, Optimization, Recommendations, etc.) providing autonomous, contextually aware synthetic decisioning that emulates human capabilities, is posed to have a transformative impact on service providers. Customers can expect faster, more relevant interactions with copilots. Service providers will increasingly use AI to offer workers more flexibility, to support and empower hitherto untapped new generalist talent, and achieve optimum efficiency with dynamic planning, scheduling and optimization.
Autonomously curated knowledge articles will continually update themselves. AI agents for assets will provide new revenue for service providers, reducing service costs and increasing customer loyalty. AI is set to drive a revolution in service, benefiting all stakeholders and my future posts through this year will comment on these, and other, emerging trends. li
To find out more about the ways IFS is already leveraging AI to manage and improve service, and to understand why Industrial ai is IFS.ai visit IFS.ai.
