Predicting Building Operations Risks With Data

Are your alarms already too late? Most teams know the feeling. A system runs quietly until one day it does not, then the work turns urgent fast. The problem often starts much earlier. Traditional BAS alarms wait for a hard threshold, so the signal appears only after the issue has already grown into a disruption. Condition-aware operations move that timeline forward. With AI applied to BAS trend data, teams can watch for small changes in runtime, cycling, efficiency, or performance against historical patterns. These shifts may be too subtle to trigger an alarm, but they can still be the first measurable signs of trouble. Take a chilled water pump. In a reactive model, the response begins after failure, reduced performance, or a comfort complaint. In a condition-aware model, a gradual rise in energy draw relative to output can be flagged earlier as a developing risk. As the pattern continues, the system can raise confidence over time, giving the team evidence to plan around instead of a single event to chase. That changes the work. Parts can be ordered before the window closes. Service can be scheduled when the building can absorb it. Teams spend less time guessing when to act and more time deciding based on a clear trend. Practical Next Steps: • Review weak signals - Look for assets where trend data shows small repeat changes before alarms ever fire. • Prioritize planned windows - Use early anomaly detection to shift service into scheduled maintenance time. • Build evidence over time - Treat emerging patterns as decision support, not just one-off alerts. • Focus on business impact - Fewer emergency callouts can mean less disruption across tenants and operations. The real value of condition-aware operations is giving teams more time to act before small issues become expensive emergencies. #BuildingAutomation, #FacilityManagement, #PredictiveMaintenance, #OperationalExcellence, #AI, #RefreshWithRyza, #NodeWork

Most asset failures are avoidable when risks are systematically identified and managed. After years of working with industrial facilities, I've found that effective risk management requires mastering five complementary frameworks: 1) HAZOP/HAZID: The foundation of process safety • HAZID provides early, broad-brush hazard identification • HAZOP deliversa systematic analysis of process deviations • Digital transformation now allows these assessments to feed directly into maintenance systems 2) FMEA (Failure Modes and Effects Analysis) • The comprehensive failure analysis framework • Now enhanced through digital twins that can simulate thousands of potential scenarios • Predictive models identify vulnerabilities that would be impossible to spot manually 3) CRA (Corrosion Risk Assessment) • Specialized analysis for material degradation mechanisms • Modern distributed sensing networks detect moisture ingress and corrosion in real-time • Early detection means addressing issues months before traditional methods would find them 4) RBI (Risk-Based Inspection) • The intelligence layer that optimizes inspection resources • AI algorithms now continuously recalculate priorities as conditions change • No more relying on outdated static schedules or calendar-based inspections 5) IOW (Integrity Operating Windows) • Defines the safe operational limits for process variables • Real-time monitoring ensures operations stay within these boundaries • Automatic alerts when parameters approach critical thresholds The power comes from integration. One refinery I worked with linked all five frameworks through a unified digital platform. Their system automatically flags when operating conditions might trigger corrosion mechanisms identified in their CRA, then updates inspection priorities in real-time. Is your organization still managing these as separate activities, or have you begun integrating them into a cohesive digital risk management strategy? *** P.S.: Looking for more in-depth industrial insights? Follow me for more on Industry 4.0, Predictive Maintenance, and the future of Corrosion Monitoring.

How to Quantify Risk: Turning Uncertainty into Insight In risk management, quantification is where strategy meets science. Qualitative assessments help identify and describe risks, but quantification is what turns these insights into actionable intelligence. So how do you quantify risk? 1. Use the Formula: Risk = Probability × Impact At its core, risk quantification involves multiplying the likelihood of an event by the financial or operational impact if it occurs. For example: A data breach that has a 10% chance of happening and could cost $1 million in damages results in a quantified risk of $100,000. 2. Apply Scenario Analysis Define a range of plausible outcomes—best case, worst case, and most likely—and assign probabilities to each. This allows you to: • Prepare for tail risks • Understand potential volatility in financial results 3. Use Monte Carlo Simulations These simulate thousands of outcomes by applying random values to input variables. It’s especially powerful for complex, interrelated risks like those in finance, investments, or supply chains. 4. Leverage Data Analysis for Pattern Detection Data is the lifeblood of modern risk management. Through historical trend analysis, time series modeling, and correlation studies, we can detect weak signals and emerging threats. Accurate data allows you to: • Track exposure over time • Benchmark risks across departments or industries • Continuously refine models with real-world feedback 5. Integrate AI for Predictive Insights Artificial Intelligence (AI) is reshaping how we measure and manage risk. Machine learning algorithms can: • Detect anomalies in real time • Predict future losses based on past behaviors • Automate risk scoring and escalation AI not only increases accuracy but also reduces manual effort and bias, allowing teams to focus on decision-making rather than data wrangling. 6. Build Risk Matrices with Numerical Scales Rather than using “Low-Medium-High,” assign numbers to likelihood and impact (e.g., 1–5 scale). This helps: • Rank risks objectively • Identify those that need immediate attention 7. Track Key Risk Indicators (KRIs) KRIs provide measurable signals of increasing or decreasing risk exposure. Examples include: • Rising customer complaint rates = Reputational risk • High turnover = Operational risk • Increasing leverage = Financial risk ⸻ Why it Matters Quantifying risk allows organizations to prioritize effectively, allocate resources wisely, and justify strategic decisions to stakeholders and regulators. In an era where uncertainty is the new normal, those who combine data analysis, AI, and quantitative tools will lead the way. #RiskManagement #QuantitativeRisk #ERM #AIinRisk #DataDriven #ScenarioAnalysis #MonteCarlo #FinanceLeadership #KRI #PredictiveAnalytics #ArtificialIntelligence

What if the fastest way to cut outages and water loss isn't more steel but more signal? When 240,000 mains break in the U.S. each year and ~2.1 trillion gallons are wasted, do we really have a pipe problem, or a data problem? My work sits at the intersection of utility ops and data. Drawing on peer-reviewed studies and sector pilots, here's what the evidence shows. Aging networks, non-revenue water (NRW) >30–40% in many systems, and thin O&M budgets keep utilities stuck in reactive mode: fixing bursts, not preventing them. But the good news is AI is already shifting utilities to predictive maintenance, real-time anomaly detection, and smarter operations. Here are 5 examples of how AI is already cutting losses and extending asset life: 1. Predictive main-break risk ranking (likelihood × consequence) Tucson's ML model ingests 12+ years of breaks plus soil, climate, and land-use to assign per-pipe risk. Engineers target the top-risk segments first, moving from age-based replacement to risk-based renewal. 2. Acoustic + ML leak hunting at network scale A U.S. Southeast city instrumented ~70 miles of at-risk pipe. AI flagged 50 hidden leaks (two ≈10 gpm mains), enabling repairs before bursts. Total saved ≈167 million gallons/year, and the same dataset reprioritized future renewals toward the weakest corridors. 3. Cutting non-revenue water with AI triage In Arizona, an AI leak-detection platform helped drive NRW from ~27% → ~10% by ranking leak likelihood/severity, focusing night-flow patrols, and shrinking time-to-repair, recovering revenue while reducing pressure shocks. 4. Energy and process optimization in treatment Aeration can be up to ~60% of plant energy. AI controllers tune dissolved oxygen (DO) setpoints and blower speeds to match real-time load, maintaining effluent quality while cutting energy per cubic meter (kWh/m³) and chemical over-dosing, and extending asset life. 5. Quality anomaly detection: catch it before customers do ML watches turbidity, chlorine, pH, and spectral signals and flags off-normal patterns (e.g., algal bloom signatures, intrusion risk). Operators get early alerts to adjust treatment or isolate zones—turning hours-late lab surprises into minutes-fast responses. While replacing pipes and upgrading SCADA is often the default path to reliability, it's not the only way. Key takeaway: Start with an AI-readiness pilot, not a moonshot. Instrument one critical zone, unify SCADA + work orders + GIS, and pick 2–3 KPIs tied to your biggest pain point: breaks/100 km, NRW %, energy per cubic meter (kWh/m³), mean time-to-repair, or leak volume avoided. (E.g., if NRW is bleeding revenue, track NRW % + leak volume avoided.) If the pilot doesn't move them in 90 days, recalibrate or stop. Where would AI pay back fastest in your system today: break prevention, NRW, energy, or water-quality compliance? Drop your baseline metric and I'll suggest a pilot scope. Repost to help your network. Follow Yulia Titova for more water insights.

The stack of paper of documentation for a jobsite is 3x higher than the building itself? Silicon Valley didn't realize that they actually invented RAG for us... 🏗️ Every day, thousands of decisions are made on a job site. Schedules shift, materials move, safety incidents get logged, and teams problem-solve in real time. Behind all of this? A mountain of unstructured data—RFIs, emails, daily reports, meeting notes, change orders, messages. The challenge? Critical insights are buried in these scattered documents and conversations. The industry has always relied on gut instinct and experience to navigate complexity—but what if we could actually use all this data to work smarter, move faster, and let builders build? This is where AI changes the game. Instead of manually digging through reports or reacting after problems surface, AI can connect the dots across all that unstructured data—unlocking real-time insights that drive impact. 🚧 Superintendents can get instant visibility into recurring safety risks before incidents happen. 📊 Project managers can pinpoint common causes of delays and adjust plans proactively. 💡 Executives can ask, “What factors are driving cost overruns across our projects?”—and get a clear, data-backed answer. At Trunk Tools, we’re laser-focused on bringing these capabilities to the field, so construction pros can spend less time hunting for information and more time doing what they do best—building. AI isn’t about replacing expertise—it’s about amplifying it. The future of construction isn’t just about breaking ground—it’s about breaking barriers in how we use data to build smarter, safer, and more efficiently. And we’re just getting started. 🚀🏗️ #ConstructionTech #AI #Innovation #TrunkTools #LetBuildersBuild

⚠️What if the Lesson Learned could be use to predict future results using Monte Carlo simulations & ML? Every shutdown we face carries its own set of risks compressed schedules, scope gaps, contractor challenges. For those of us who live and breathe turnaround planning, we know how quickly small oversights can snowball into cost overruns or delays. During the past months, as part of my MSc Final Project, I’ve been researching and working on the Predictive Lessons Learned (PLL) Model a structured way to transform past experiences into forward-looking risk predictions. Instead of just storing “lessons learned” in a database, PLL applies Bayesian smoothing, odds ratios, and logistic modeling to convert those lessons into measurable risk probabilities. ✅ What does this mean for professionals in the field? It means we can move beyond intuition and “gut feel,” and actually quantify: 📌How much contractor inexperience increases schedule risk? 📌What compressed timelines really do to probabilities of overrun. 📌Which factors matter most, so we can act before risks materialize. 📊The model is simple enough to run in Excel yet rigorous enough to expand with machine learning and explainable AI. Most importantly it gives turnaround teams transparency and clarity when it matters most: in the heat of decision-making. At the end of the day, no model eliminates uncertainty. But having a data-informed, explainable framework empowers us to manage it making turnarounds safer, smarter, and more predictable. 👉 To my turnaround peers: How are you currently capturing and using lessons learned? Are they driving predictive insights or just sitting in archives? #TurnaroundManagement #RiskManagement #LessonsLearned #PredictiveAnalytics #OilAndGas #PlantMaintenance

Most organizations are still treating AI, Digital Twins, and AIOps as separate initiatives. The real value appears when they work together. A digital twin gives you a real-time model of your facility. AIOps continuously analyzes telemetry, identifies patterns, predicts failures, and recommends or automates corrective actions. Together, they shift operations from reactive to predictive. Where are organizations seeing the biggest impact? ✅ Cooling Optimization Digital twins combined with AIOps can continuously optimize cooling setpoints, water temperatures, flow rates, and airflow. The result is lower energy consumption, increased capacity, and fewer thermal events. ✅ Predictive Maintenance Instead of waiting for a pump, CDU, UPS, generator, or chiller to fail, AI models identify abnormal behavior before it becomes an outage. Maintenance becomes planned instead of emergency-driven. ✅ Faster Commissioning & Change Management Teams can validate sequences of operation, interlocks, and failure scenarios in a virtual environment before touching production systems. This reduces commissioning cycles, improves quality, and lowers operational risk. The lesson is simple: The future of facility operations is not just more sensors, more dashboards, or more AI. It's creating a digital representation of your environment and using intelligence to continuously optimize performance, reliability, and efficiency. For data centers supporting AI workloads, where power density and cooling demands continue to rise, this approach is quickly becoming a competitive advantage rather than an innovation project. The organizations that build these capabilities now will be the ones operating more efficiently, scaling faster, and avoiding the costly surprises that traditional operations teams spend their days reacting to. #DataCenter #AIOps #DigitalTwin #ArtificialIntelligence #DataCenterOperations #InfrastructureManagement #PredictiveMaintenance #FacilityManagement #CriticalInfrastructure #ITOperations #DigitalTransformation #OperationalExcellence #AIInfrastructure #MissionCritical #FutureOfWork

In a post earlier this week I outlined the problem with today’s physical climate risk models. Essentially, they boil down building vulnerability to a single loss number, primarily stemming from collected claims data — a top-down approach that doesn't provide asset-level traceability. But buildings are comprised of hundreds or thousands of components and systems — mechanical, electrical, plumbing, architectural, and structural elements — many of which are interdependent. That’s why we need to model building vulnerability and physical damage at the building component level, using a bottom-up, first principles of engineering-based approach. With component-based modeling, we can trace: • the hazard intensity each component is likely to experience • the likelihood a component is damaged, and to what severity • what specific repair actions would be required • how those repairs translate into time and cost With this type of approach to vulnerability modeling, we can explain so much more: Why is the risk high? What’s driving the loss? And what specific interventions (and I don't mean flood walls!) would change the outcome? When you can connect hazard → component damage → repair actions → business impact, the analysis becomes technically defensible — and stakeholders can make retrofit, acquisition, design, and resilience decisions with far more confidence. Importantly, this approach to quantifying credible asset-level behavior scales naturally to portfolios of hundreds or thousands of buildings.

Ever feel like you're missing pieces of the puzzle when it comes to predicting system performance? Physical sensors are invaluable, but they can't tell us everything that's happening inside our complex designs or where issues might arise in the future.     My experience in digital transformation has taught me that true operational excellence comes from seeing beyond the obvious. It's about bridging the gap where physical data ends and deeper insight begins. We often face situations where critical temperatures, pressures, or erosion rates are needed in locations without sensors, or we need to understand future events that today's data simply can't capture.     That's where the power of virtual sensing, powered by predictive engineering analytics, really shines. Imagine simulating any real-world physical behavior from fluid mechanics to heat transfer to get a complete picture of your system. This isn't just about design; it's about embedding this predictive capability into the operational digital twin.     Take, for example, a heat exchanger. Sensors might flag a high temperature, but simulation reveals the precise flow distribution causing those temperature gradients and the resulting stresses. Or in subsea production, where understanding thermal performance is critical for hydrate avoidance. While high-fidelity simulations are great for design, system-level simulations, tuned by that detailed data, provide the real-time insights we need for operations.     This approach transforms raw field data into actionable engineering judgment. It means extending maintenance schedules with confidence, understanding system capacity beyond design conditions, and making proactive decisions that optimize performance and ensure integrity for years.     What challenges are you facing in gaining full visibility into your system's performance? How could predictive analytics unlock new possibilities for your operations? I'd love to hear your thoughts. 

Last week I met with the head of automation & digitalization at a large mining company. Their ask was simple (and incredibly common): “How do we get an early indication that a process is about to fail, trip, or drift out of yield/quality—before an alarm even fires?” That question quickly shifted from “what if” to “when can we do it?”—because with predictive analytics that combines real-time and historical data, it’s possible to move from reactive operations to proactive, preventative action (and do it on top of existing systems—without ripping and replacing the control platform). That’s exactly why I’m excited about the commercial launch of Honeywell Experion Operations Assistant (part of the Experion PKS ecosystem): it brings AI-powered decision support and predictive intelligence into the control room—helping operators anticipate unsafe conditions and production losses before they escalate. Predicts and flags emerging issues earlier—in the pilot, predictions were made an average of 5–10 minutes before alarm incidents would have happened. Merges real-time operational insights with historical context to support faster, more confident decisions. Designed to integrate into existing control room environments, leveraging legacy and site-specific data. Supports productivity and safety goals by reducing unplanned downtime and avoiding potential events. If you’re responsible for operations, automation, or digital transformation, I’m curious: what’s the earliest signal you wish your control room had—before alarms? More details here (press release): https://lnkd.in/gXfkC5dt #ProcessAutomation #Mining #ControlRoom #PredictiveAnalytics #IndustrialAI #DigitalTransformation #OperationalExcellence #Safety