How Johnson Controls Transformed Service-Parts Planning with Agentic Intelligence
JCI enhances service-part forecast accuracy, part availability, and execution discipline by integrating adaptive, agent-driven planning into daily operations.
Visión general
Johnson Controls International (JCI) is a global leader in smart, healthy, and sustainable building solutions, supporting customers across commercial, industrial, and institutional markets worldwide. A critical component of JCI’s service business is its global service-parts supply chain, which supports uptime, maintenance, and repair across a vast installed base.
JCI manages thousands of service parts characterized by low-volume, intermittent demand, long product lifecycles, and multiple demand sources—including maintenance, repair, warranty, and one-time or custom orders. Maintaining high part availability while controlling inventory and cost is essential to service performance and customer satisfaction.
To address these challenges, JCI launched a focused Kaizen initiative aimed at improving forecast accuracy as a primary lever to increase part availability, reduce backorders, and improve execution across its service network.
Desafío
Despite its scale, JCI’s service-parts planning process relied heavily on manual intervention. Demand planners faced an overwhelming volume of alerts and exceptions that limited their ability to focus on high-value decisions.
Key challenges included:
- More than 12,000 forecast and demand alerts, far exceeding what planners could realistically manage
- Low and lumpy demand patterns that reduced the effectiveness of traditional forecasting methods
- Limited visibility into demand type (maintenance, repair, warranty, or one-time orders), requiring manual investigation
- Forecast adjustments driven by spreadsheets, phone calls, and tribal knowledge
- No formal, integrated SIOP (Sales, Inventory, and Operations Planning) process
As a result, planners spent weeks each month cleansing demand and reacting to alerts, with limited impact on part availability, OTIF, and backorder age.
¿Está listo para revolucionar la planificación de su cadena de suministro?
Descubra cómo la solución de Planificación Adaptativa de la Cadena de Suministro PolymatiQ™ de ketteQ puede ayudar a su empresa a navegar de forma proactiva ante las interrupciones y optimizar el rendimiento. Para obtener más información, visite ketteQ.com.
Solución
JCI partnered with ketteQ during its Kaizen initiative to redesign the process of forecasting and planning decisions, shifting from manual, alert-driven processes to automated, agent-driven decision intelligence.
By deploying ketteQ’s adaptive planning and agentic capabilities, JCI addressed three core areas:
- Automated forecast correction and alert reduction:
Planning agents identified under- and over-forecast conditions by comparing future forecasts to historical demand, executing decision trees to adjust forecasts and resolve alerts automatically. - Demand-type classification and cleansing:
Demand was automatically categorized, allowing one-time and atypical events to be excluded from future forecasts without manual analysis. - Introduction of a formal SIOP process:
A lightweight SIOP framework aligned product management, service, and planning teams around shared dashboards and metrics.
Rather than replacing planners, ketteQ automated routine decisions within defined guardrails, allowing planners to focus on exceptions, policy decisions, and cross-functional alignment.
Resultados
JCI’s Kaizen initiative, powered by ketteQ, delivered rapid, measurable improvements across forecast accuracy, part availability, and operational efficiency:
- Forecast accuracy improved by up to 25 points, driven by:
- A 15-point improvement from agent-driven under- and over-forecast correction
- Incremental gains from machine-learning forecasting and SIOP alignment
- Forecast accuracy expected to increase from 46% to ~75%, exceeding typical industry benchmarks
- 48% reduction in planning process steps, reducing handoffs from 29 to 15
- Part availability increased from 91% to 98%, supporting the goal of fulfilling 98% of parts within two days
- Service performance improvements, including:
- OTIF Chiller (In-Network) improvement from 75% to 93%
- OTD Controls (In-Network) improvement from 42% to 95%
- Backorder age reduction from 61 days to 14 days
- ~80% of forecast alerts automatically resolved, dramatically improving planner productivity
These results demonstrated a direct link between improved forecast accuracy and improved service outcomes.
Por qué es importante
JCI’s experience shows that service-parts planning at scale cannot rely on manual processes and static forecasts. In environments characterized by intermittent demand and high service criticality, planners require systems that continuously interpret signals and act within predefined guardrails.
By embedding adaptive, agent-driven planning into its operations, JCI reduced friction, improved execution discipline, and scaled decision-making without increasing complexity.
De cara al futuro
With a global deployment underway, JCI is building on this foundation to further modernize its global service-parts planning operations. Adaptive, agent-driven decision intelligence has become a core capability enabling improved service performance while freeing planners to focus on what matters most
Otros casos prácticos
Cómo la planificación adaptativa impulsó el éxito del líder en recambios para restaurantes.
.svg){kind=logo}
El líder en equipos de calefacción, ventilación y aire acondicionado y sistemas de control consigue un aumento del 10% en las entregas a tiempo y una mejor gestión del inventario.