AI in Maximo 2026: From Condition Insight to Agentic Workflows

AI in Maximo 2026: From Condition Insight to Agentic WorkflowsMaximo Application Suite 9.x ships with AI capabilities that are operationally grounded, integrated into the daily workflow, and proven in production at a scale that would have seemed ambitious just two years ago. The release of Maximo Application Suite 9.2 in June 2026 is the most consequential step in that transition. IBM has embedded AI more deeply into the daily work of managing assets, teams, safety, and operations.

The release expands AI across reliability insights, field execution, safety and compliance workflows, document-based information extraction, and orchestration across systems. It also introduces agentic workflows designed to help guide decisions and move work forward in operationally grounded ways. Rather than treating AI as a separate layer that sits on top of Maximo, MAS 9.2 applies it where work happens, helping teams identify issues earlier, make better decisions faster, and execute with greater consistency in the field.

This article walks through the AI capabilities in Maximo as of mid-2026: Condition Insight, Visual Inspection, Maximo Assistant, the watsonx foundation, and the agentic shift that defines the 9.2 release. For each capability, we examine what it does, how it works, what production results look like, and what it means for the daily work of reliability engineers, maintenance managers, and field technicians.

Maximo Condition Insight: The Flagship AI Capability

The single most important AI capability in MAS 9.x is Maximo Condition Insight, which IBM introduced in late 2025 and which has matured into a flagship feature by mid-2026. Condition Insight is an agentic AI capability within Maximo Asset Performance Management (APM) that interprets asset data to explain asset condition, highlight emerging trends, and recommend corrective actions. It works in concert with the other MAS applications to deliver a unified, condition-based maintenance approach across the Maximo ecosystem.

Powered by IBM watsonx, Condition Insight evaluates work orders, inspections, metrics, time-series data, meter readings, Failure Mode and Effects Analysis (FMEA), and alerts to evaluate the asset's condition, uncover performance patterns, and provide clear, actionable recommendations. All of this is communicated in plain, understandable language, which is a critical design decision. The AI does not output a probability score or a confidence interval. It outputs a narrative summary that a maintenance planner can read and act on.

The architecture of Condition Insight follows a five-step agentic loop:

Condition Insight Agentic Loop:

1. GATHER CONTEXT
2. - User input (e.g., "Why is pump 1042 degrading?")
3. - System state (work orders, inspections, meter readings)
4. - Asset history (failure codes, PM records, FMEA data)

1. PLAN
2. - Determine what analysis is needed
3. - Select relevant data sources
4. - Identify evaluation steps

1. EVALUATE
2. - Apply AI models to interpret data
3. - Compare patterns against historical baselines
4. - Cross-reference with FMEA and reliability strategies

1. ACT
2. - Generate condition summary in plain language
3. - Recommend corrective actions
4. - Create or update work orders if warranted

1. OBSERVE
2. - Monitor the outcome of recommended actions
3. - Update the asset's condition profile
4. - Feed results back into the planning loop
5. `

The practical impact is significant. Before Condition Insight, a reliability engineer investigating an asset anomaly needed to manually pull work order history, review inspection results, check meter readings, cross-reference with FMEA data, and synthesize the findings into a recommendation. This process could take hours per asset, and in organizations with thousands of critical assets, it meant that only the highest-priority assets received this level of analysis. Condition Insight performs this synthesis in seconds, returning a clear, explainable summary of condition, trends, and recommended actions. This makes AI practical for every maintenance team, not just teams with dedicated data scientists.

Condition Insight removes the barrier between asset data and actionable intelligence. The combination of Maximo's trusted asset data model with watsonx AI delivers explainable, enterprise-grade AI that integrates directly into maintenance workflows. The AI does not replace the reliability engineer. It does the data gathering and initial synthesis, allowing the engineer to focus on decision-making and execution. The recommendations are transparent and explainable: the engineer can see which data points informed the analysis, what patterns were identified, and how the recommendation was derived. This transparency is essential for regulated industries where AI recommendations must be auditable.

Maximo Visual Inspection: AI at the Edge

Maximo Visual Inspection enables AI-based visual inspection with local inference directly on the device. This is not a cloud-based image analysis service. It runs on the hardware that inspectors carry into the field, which means it works in environments with no connectivity and returns results in real time without latency.

The capability uses computer vision models trained to identify defects, anomalies, and wear patterns in asset images. A technician points a mobile device camera at a piece of equipment, and the model identifies whether the asset shows signs of degradation that warrant further inspection or maintenance action. The model can be trained on custom image datasets for specific asset types, making it applicable across industries from manufacturing to utilities to transportation.

# Example: Visual Inspection workflow
# 1. Technician captures image of asset
image = capture_asset_image(asset_num="PUMP-1042")

The local inference capability is a critical differentiator. In environments like refineries, mines, and remote utilities infrastructure, network connectivity may be unreliable or unavailable. A cloud-based inspection service that requires uploading images and waiting for analysis would be impractical. Local inference means the technician gets an answer immediately, can take action on the spot, and syncs the results when connectivity is restored. The inspection results are then stored in Maximo alongside other asset data, creating a complete condition history that includes both sensor-based monitoring and visual assessment data.

Melbourne Water used IBM Maximo Visual Inspection to help combat extreme weather conditions, demonstrating the capability in a real-world utilities environment. The deployment enabled inspectors to assess infrastructure condition more rapidly and consistently, supporting the organization's resilience strategy for dealing with climate-related challenges. Boston Dynamics created a robotic dog that reads sensors and collects data that are then analyzed by IBM Maximo, further extending the visual inspection capability into automated inspection workflows that do not require a human technician on site.

Maximo Assistant on Mobile: Natural Language in the Field

Maximo Assistant on Mobile helps technicians use natural language to find asset information, review history, and complete work efficiently in the field. The assistant is integrated into the Maximo Mobile application, which means it is part of the workflow that technicians already use, not a separate tool that requires context switching.

A technician can ask the assistant questions like "What is the maintenance history for this asset?" or "What safety procedures apply to this work order?" and receive answers synthesized from Maximo data in natural language. The assistant understands the context of the current work order, the asset being worked on, and the technician's location, enabling it to provide relevant information without the technician needing to navigate through multiple screens.

Technician: "What work orders are open for this asset?"

Assistant: "There are 3 open work orders for PUMP-1042: - WO 10234: Bearing replacement (Approved, scheduled for Aug 5) - WO 10245: Monthly inspection (Waiting approval) - WO 10251: Seal leakage repair (Approved, priority High)

The most recent failure was on July 15, involving a bearing failure. The asset has been flagged for condition monitoring with a Condition Insight alert showing increased vibration readings over the past 14 days."

Technician: "Create a follow-up work order for the vibration issue."

Assistant: "Created work order WO 10278: - Asset: PUMP-1042 - Type: Corrective Maintenance - Description: Investigate vibration increase per Condition Insight alert - Priority: Medium - Status: Waiting Approval The work order has been linked to the existing Condition Insight alert." ```

The assistant reduces the time technicians spend on administrative tasks (looking up asset history, reviewing procedures, creating work orders) and increases the time they spend on actual maintenance work. It also reduces errors, because the assistant handles data entry through natural language rather than requiring technicians to navigate complex forms on mobile devices. For new technicians who may not be familiar with Maximo's interface, the assistant provides a guided experience that helps them find information and complete tasks without extensive training. This is particularly valuable for organizations with high technician turnover or contractors who need to be productive quickly.

The watsonx Foundation

All AI capabilities in MAS 9.x are powered by IBM watsonx, the enterprise AI platform that provides the underlying machine learning, natural language processing, and model serving infrastructure. The choice of watsonx as the AI foundation is significant for three reasons.

First, watsonx is designed for enterprise deployments with the security, governance, and compliance controls that asset-intensive industries require. AI models running on watsonx operate within a governance framework that tracks model provenance, monitors for bias and drift, and provides audit trails for regulatory compliance. This is not consumer AI running in a black box. It is enterprise AI with the controls that regulated industries demand.

Second, watsonx supports both IBM-developed models and third-party models. This means organizations are not locked into a single AI provider. If a specific use case is better served by a specialized model from another provider, that model can be integrated through the watsonx platform. The Maximo Predict application already supports this flexibility: data scientists can build and train custom predictive models using default notebooks or completely custom notebooks, as long as the models are deployed in Watson Machine Learning.

Third, watsonx enables organizations to use their own data to train and fine-tune models. Maximo Predict allows data scientists to create groups of assets and generate different types of predictions for those assets, such as current failure probability or estimated failure dates. The models can be retrained as new data becomes available, improving prediction accuracy over time. The predictions populate a Predictions section for each asset, and work queues can track assets with high failure probability or assets predicted to fail before the next scheduled preventive maintenance work order.

# Example: Maximo Predict workflow for custom model

The deployment flexibility of the AI Service in MAS 9.2 further extends the options. Three deployment patterns are available: Full SaaS (no new purchase required, uses existing AppPoints), Hybrid AI SaaS (SaaS AppPoints required, no new infrastructure), and Full Customer Managed (includes a watsonx.ai license for use with AI Service only). This range allows organizations to choose the deployment model that fits their cloud strategy, data residency requirements, and existing infrastructure investments.

The Agentic Shift in MAS 9.2

The 9.2 release introduces what IBM calls agentic workflows, which represent a fundamental shift in how AI participates in Maximo operations. Previous AI capabilities in Maximo were primarily analytical: they analyzed data, identified patterns, and presented insights for human decision-makers to act on. Agentic workflows go further. They enable AI to participate in the execution of work, guiding decisions and moving work forward across processes in practical, operationally grounded ways.

The MCP Server, discussed in detail in the integration architecture article in this series, is the technical foundation for this agentic shift. It provides the standardized protocol through which AI agents interact with Maximo Manage APIs, enabling them to create work orders, query assets, update statuses, and perform other operational tasks within the governance framework of the platform.

The combination of Condition Insight (which provides the analytical foundation), the MCP Server (which provides the integration mechanism), and agentic workflows (which provide the execution model) creates a new paradigm for AI in asset management. AI is no longer just a dashboard that tells you what might fail. It is a participant in the workflow that can recommend actions, create work orders, and track the outcome of interventions, all within the governance and security framework of the Maximo platform.

This does not mean AI replaces human decision-making. Every agentic workflow includes human checkpoints where the recommendations are reviewed and approved before execution. The AI does the heavy lifting of data gathering, analysis, and initial recommendation. The human reviews, approves, and oversees the execution. This is the model that regulated industries require, and it is the model that MAS 9.2 implements. The agentic capabilities also extend to document-based information extraction, where AI can parse maintenance manuals, safety procedures, and compliance documents to extract structured information that feeds into Maximo workflows, reducing the manual effort required to digitize paper-based processes.

Practical Implications

For reliability engineers, the AI capabilities in MAS 9.2 change the daily workflow in two ways. First, Condition Insight eliminates the data gathering and synthesis phase of asset investigation. Instead of spending hours pulling work order history, reviewing inspections, and cross-referencing FMEA data, the engineer starts with a synthesized condition summary and recommended actions. The time saved can be redirected toward higher-value activities like reliability strategy development and root cause analysis. Second, the continuous monitoring enabled by Condition Insight means that the engineer's attention is directed to assets that need it, rather than being spread across the entire portfolio based on calendar schedules.

For maintenance managers, the combination of Condition Insight and Maximo Predict provides a continuous risk assessment of the asset portfolio. Assets are no longer evaluated only during periodic reliability reviews. They are continuously monitored, with risk assessments updated as new data arrives. Work queues automatically surface high-risk assets, enabling proactive intervention before failures occur. This shifts the maintenance organization from a reactive posture (responding to failures) to a proactive posture (intervening before failures happen), which reduces unplanned downtime, extends asset useful life, and optimizes maintenance spending.

For field technicians, Maximo Assistant and Visual Inspection reduce the administrative burden of field work. The assistant provides instant access to asset information and procedures through natural language, eliminating the need to navigate complex application interfaces on mobile devices. Visual Inspection provides instant defect identification, enabling technicians to make immediate decisions about whether an asset needs further attention. Together, these capabilities increase first-time fix rates and reduce the time spent per work order.

For IT leaders, the watsonx foundation provides the security, governance, and flexibility that enterprise AI deployments require. The ability to use both IBM and third-party models, to train custom models on organizational data, and to deploy within a governed framework makes the AI investment defensible across regulatory environments. The agentic workflows introduced in 9.2 are the beginning, not the end, of this shift. Organizations that establish the foundation now will be positioned to adopt increasingly sophisticated AI capabilities as they are released in future MAS updates.

Bottom Line

AI in Maximo has moved from experimental to operational. Condition Insight delivers agentic asset condition analysis in seconds that would take a reliability engineer hours to perform manually. Visual Inspection brings AI defect detection to the field with local inference on mobile devices. Maximo Assistant gives technicians natural language access to asset information and work order management. The watsonx foundation provides the enterprise-grade AI infrastructure that regulated industries require. And the agentic workflows in MAS 9.2 represent the beginning of a new paradigm where AI participates in asset management workflows within a governed, auditable framework. For organizations on MAS, the AI capabilities are not a future roadmap. They are available now, integrated into the daily work, and ready for production deployment.