Using Analytics to Measure Productivity

Real-time processing is becoming the defining capability that separates operational 3D systems from research demos. Not just speed - though that matters. But the entire philosophy of how you architect 3D workflows. Traditional approach: capture → transfer → process → analyze → report. Timeline: days to weeks. Real-time approach: capture → stream → process on edge → continuous analysis → instant alerts. The difference isn't just faster results. It's fundamentally different decision-making. When analysis happens in real-time, you can adjust construction activities while equipment is still on-site. Catch errors while crews are present. Validate progress before moving to the next phase. When analysis happens days later, you're always reacting to old information. Here's what I'm seeing emerge: distributed 3D processing where compute happens at the edge, not in the cloud. LiDAR sensors with built-in processing. Drones that segment point clouds in-flight. Site tablets that run change detection locally. This matters for construction sites with limited connectivity. For workflows that need instant feedback. For systems that must work without cloud dependency. Two tutorials that show you how to build these automated processing pipelines: → How to Automate LiDAR Point Cloud Processing with Python (13 min) https://lnkd.in/eBiZvRhU → 3D Python Workflows for LiDAR City Models - A Step-by-Step Guide (38 min) https://lnkd.in/egYfcCrq 51 minutes total. Automated preprocessing pipelines and multi-modal 3D workflows at scale. But here's where this gets really interesting: the convergence of real-time processing and predictive analytics. Your processing pipeline doesn't just analyze current scans. It compares them to historical patterns. Predicts where deviations are likely. Prioritizes inspection zones automatically. Construction has massive amounts of data but very little real-time intelligence. The teams building real-time systems now will define how sites operate in the next decade.

🚀 Real-Time AI for Smart Manufacturing — Liquid Gel Bottle Filling Monitoring System Excited to share a project I’ve been working on: a computer vision pipeline for monitoring liquid gel bottle production lines in real time. 💡 What it does: • Detects and classifies bottles into Empty, Filling, and Filled. • Tracks each bottle with a unique ID across frames • Outputs an annotated video with live production statistics 🧠 Key Technologies & Innovations: 🔹 Ultralytics YOLO26 Object Detection Custom-trained model for high-accuracy detection of bottle states(Roboflow annotations) with optimized inference thresholds. 🔹 Deep SORT Tracking Ensures each bottle is tracked consistently, enabling reliable counting without duplication. 🔹 Smart Counting Logic Each bottle is counted only once using track IDs — ensuring accurate production metrics. 🔹 CUDA Acceleration ⚡ GPU-powered inference (FP16 + optimized input size) for real-time performance. 🔹 Threaded Video Processing Separates frame loading from inference to eliminate bottlenecks. 🔹 Custom Visualization Layer Color-coded bounding boxes Transparent overlays Clean labeling system 🔹 Live Donut Chart 📊 Real-time visualization of production distribution — rendered directly with OpenCV. ⚙️ Performance Highlights: • Smooth real-time processing • Optimized memory & GPU usage • Dual-resolution output (high-quality recording + consistent display) 📁 Modular Pipeline Versions: • CPU baseline • CUDA-accelerated • High-performance optimized • Full version with live analytics dashboard 🎯 Why it matters: This system demonstrates how AI + Computer Vision can bring visibility, efficiency, and intelligence to industrial production lines — a key step toward Industry 4.0. 🎯 Key Takeaways • Combining detection + tracking is essential for reliable counting • System-level optimizations (threading, memory reuse) matter as much as model accuracy • Avoiding external plotting libraries significantly improves real-time performance • Careful GPU utilization can turn a standard pipeline into a production-ready system #OpenToWork #AIEngineer #ComputerVisionEngineer #MachineLearningEngineer #SoftwareEngineer #DeepLearning #ComputerVision #MLOps #Python #OpenCV #RealTimeSystems #EdgeAI #AIProjects #TechCareers #ComputerVision #DeepLearning #AI #ObjectDetection #YOLO #MultiObjectTracking #EdgeAI #RealTimeSystems #OpenCV #SmartManufacturing #Industry40 #AIEngineering #TechInnovation

Managing a business with yesterday’s data is like driving while looking in the rearview mirror. A few weeks ago, I shared how we’re using AI to drive better outcomes for our partners and their merchants. But generating meaningful insights takes more than just smart tools — it requires a shift in mindset. At NMI, we’re moving from 𝘳𝘦𝘢𝘳𝘷𝘪𝘦𝘸 𝘮𝘪𝘳𝘳𝘰𝘳 𝘮𝘦𝘵𝘳𝘪𝘤𝘴 to 𝘸𝘪𝘯𝘥𝘴𝘩𝘪𝘦𝘭𝘥 𝘮𝘦𝘵𝘳𝘪𝘤𝘴: real-time signals that help us actively steer the business forward, not just analyze where we’ve been. As part of this shift, we’ve developed multi-point partner health scores that give us a holistic, dynamic view of customer health across our ecosystem. To enable this, we’ve: •Integrated analytics into our channel account dashboards (and update them monthly) •Blended signals from product usage, billing, support interactions, and customer sentiment •Invested in streaming data to spot lags in transactions and provide more consultative, timely support Real-time insights allow us to act on what we see. These insights feed into our regular partner health check-ins, and when warning signs appear, we proactively reach out to help partners course-correct. Windshield metrics not only help us manage our business more effectively, they also enable us to better support our partners. Over time, our goal is to evolve these analytics into a solution our partners can offer to their own merchants, strengthening every link in the value chain — from NMI to our partners, and from our partners to their customers. Moving towards windshield analytics is just one way we’re continuously evolving to enhance the partner experience. How does your organization approach data? Are you still operating on “rearview” insights? Or have you adopted real-time analytics? Let me know in the comments! 👇 #Fintech #Metrics #RealTimeInsights #TechLeadership #DataDrivenLeadership

A year ago, a friend working at a healthcare e-commerce startup struggled with delayed order deliveries. Despite having accurate stock levels, some customers received their orders late. The operations team blamed warehouse inefficiencies, but the real issue was hidden in the data. Investigating the Root Cause with SQL 1️⃣ Measuring Average Fulfillment Time First, we calculated the time taken from order placement to dispatch for each order. SELECT order_id, warehouse_id, DATEDIFF(minute, order_placed_time, dispatch_time) AS fulfillment_time FROM orders; 🔹 Insight: Some warehouses were consistently slower than others. 2️⃣ Identifying Bottlenecks Next, we checked if warehouse processing time was affected by order volume. SELECT warehouse_id, COUNT(order_id) AS total_orders, AVG(DATEDIFF(minute, order_placed_time, dispatch_time)) AS avg_fulfillment_time FROM orders GROUP BY warehouse_id ORDER BY avg_fulfillment_time DESC; 🔹 Insight: Warehouses with higher order volumes had longer processing times, pointing to capacity issues. 3️⃣ Detecting Delays in High-Priority Orders Urgent orders (medications) were supposed to be processed faster. We checked if they were actually prioritized. SELECT order_id, priority_level, DATEDIFF(minute, order_placed_time, dispatch_time) AS fulfillment_time FROM orders WHERE priority_level = 'High' ORDER BY fulfillment_time DESC; 🔹 Insight: High-priority orders weren’t always processed first, revealing an issue with order prioritization logic. Challenges Faced Slow Query Performance – Indexed order_placed_time and dispatch_time to speed up calculations. Identifying True Bottlenecks – Cross-checked with staffing data to confirm that delays were due to capacity, not inefficiency. Operational Resistance – Warehouse teams resisted changes, so we presented data visually to show problem areas. Business Impact ✔ 15% reduction in delivery delays after adjusting warehouse staffing. ✔ High-priority orders processed 40% faster by improving sorting logic. ✔ Data-driven decision-making enabled proactive warehouse management. Key Takeaway: SQL isn’t just about querying data—it’s about uncovering hidden inefficiencies and driving operational improvements. Have you used data to solve real-world logistics problems? Let’s discuss!

𝗗𝗼𝗻’𝘁 𝗚𝘂𝗲𝘀𝘀, 𝗗𝗶𝗮𝗴𝗻𝗼𝘀𝗲: 𝗔 𝗣𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹 𝗧𝗶𝗽 𝗳𝗼𝗿 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗢𝗽𝘁𝗶𝗺𝗶𝘇𝗮𝘁𝗶𝗼𝗻 Last week, my team almost wasted days optimizing the wrong thing. A specific search function was frustratingly slow for users under load. The gut feel was slow SQL queries - seemed like a simple fix. But if experience has taught me anything, it’s to verify first, no matter how obvious the problem looks. Here’s the 3-step process they followed to get the best result: 1️⃣ 𝗠𝗲𝗮𝘀𝘂𝗿𝗲 𝗙𝗶𝗿𝘀𝘁: 𝗨𝘀𝗲 𝗠𝗲𝘁𝗿𝗶𝗰𝘀 𝗮𝗻𝗱 𝗣𝗿𝗼𝗳𝗶𝗹𝗶𝗻𝗴 They could have spent days tweaking queries based on assumptions. Instead, we invested in proper tooling: ✅ 𝗞𝟲 simulating the requests to put the system under a similar load ✅ 𝗢𝗽𝗲𝗻𝗧𝗲𝗹𝗲𝗺𝗲𝘁𝗿𝘆 to capture traces and metrics across all backend systems ✅ 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀 as our APM to bring traces, metrics and logs into one place for analysis The surprise: CPU-bound processing was the real culprit, not the database. 2️⃣ 𝗣𝗶𝗻𝗽𝗼𝗶𝗻𝘁 𝘁𝗵𝗲 𝗕𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 With OpenTelemetry, we could trace requests end to end: ✅ API response times ✅ Business logic execution ✅ Actual SQL query performance ✅ External service calls The reality check: complex nested LINQ logic in the search algorithm was spiking CPU usage, using up more than 60% of request time. 3️⃣ 𝗙𝗶𝘅 𝘄𝗶𝘁𝗵 𝗙𝗼𝗰𝘂𝘀 Once the facts were clear, they could prioritize: ✅ First: Fix the CPU bottlenecks ✅ Next: Still improve the SQL queries and optimize table structures ✅ Throughout: Monitor each change to prove real gains 𝗧𝗵𝗲 𝗸𝗲𝘆 𝘁𝗮𝗸𝗲𝗮𝘄𝗮𝘆: performance tuning should never be guesswork. Measure, fix, then measure again. Listen to your gut, but verify first. The time spent diagnosing prevented less-effective "fixes" and saved us days. 𝘞𝘩𝘢𝘵'𝘴 𝘺𝘰𝘶𝘳 𝘨𝘰-𝘵𝘰 𝘵𝘰𝘰𝘭 𝘴𝘵𝘢𝘤𝘬 𝘧𝘰𝘳 𝘱𝘦𝘳𝘧𝘰𝘳𝘮𝘢𝘯𝘤𝘦 𝘥𝘪𝘢𝘨𝘯𝘰𝘴𝘵𝘪𝘤𝘴? 𝘈𝘯𝘺 𝘩𝘪𝘥𝘥𝘦𝘯 𝘨𝘦𝘮𝘴 𝘺𝘰𝘶'𝘥 𝘳𝘦𝘤𝘰𝘮𝘮𝘦𝘯𝘥?

Most teams think productivity problems come from people. The data says calendars are the real bottleneck. Employees reporting higher productivity average 4.4 daily focus hours. Employees reporting lower productivity average 2.7 daily focus hours. That gap is not motivation. That gap is meeting design. High meeting loads silently erase deep work. Fragmented calendars destroy momentum before work even begins. Focus time is not a perk. Focus time is a production input. Once focus drops below 3 hours, productivity perceptions collapse. Meetings do not scale linearly with outcomes. They scale linearly with distraction. Every additional meeting increases recovery time across the day. Context switching compounds faster than leaders expect. Most teams never notice the tipping point. They feel busy while output quietly slows. The chart makes this painfully obvious. Productive teams protect uninterrupted blocks. Unproductive teams optimize for constant availability. Availability feels responsive. Focus actually drives results. This is not about fewer meetings everywhere. It is about better sequencing and intent. The same meetings can produce radically different days. Intentional scheduling doubles usable focus without reducing collaboration. That is leverage most teams never pull. Leaders often ask why execution feels slower. The answer lives on the calendar. If productivity is falling, start measuring focus. If focus is low, stop blaming effort. What would change if focus hours became a leadership metric?

What they don't teach you in #NIR grad school Several times a week I get asked to present a brief introduction about NIR to people. Gradually, I have learned that people's eyes glaze over if I try and tell them how the thermostat works. There are so many interesting details. So I have had to train myself to follow the 11th Commandment: Thou shalt not be boring. So I try to capture 3 points and 3 sub points as my memory spots and add fresh details under each one for each audience so it stays fresh. You can do better than this but this is way better than trying to do it like I would have in Grad School. Here's my intro to NIR outline 1. Real-time Results Analyze materials on the line—no delays: NIR gives you immediate feedback on material composition as it’s processed, eliminating the need to send samples to a lab and wait for results. Catch variations early, prevent off-spec products: Since you’re seeing data as it happens, you can address issues before they escalate, reducing rework or product loss. Faster decisions = faster improvements: With quick data insights, you can make operational adjustments on the fly, keeping production running smoothly and efficiently. 2. Optimize Operations Consistent product quality, less waste: NIR ensures uniform quality by continuously monitoring product parameters, minimizing variations and helping to maintain tight specs. Adjust processes immediately for better yield: By using real-time data, operators can tweak parameters like moisture, fat, or protein levels in real time, maximizing the use of raw materials. Lower costs directly from reduced variability: Consistent quality and optimized yields mean less waste, translating directly into lower production costs. 3. Boost Your Bottom Line Cut expenses, improve efficiency: By improving yield and reducing rework, NIR directly lowers material costs and energy use, driving better margins. Drive savings straight to profit margins: Every dollar saved from reduced waste or better efficiency adds to your profit margins, making NIR a tool for cost-saving and profit enhancement. Bigger bonuses tied to better performance: When plant performance improves, so do your metrics, which means a stronger case for performance-based bonuses. Suggestions are welcome. I am always trying to learn a better way.

Busy plants aren’t always productive plants. That’s the fastest way to lose money quietly. Most plants look busy. Most machines look utilized. Most dashboards look green. And yet… output stalls, orders slip, and customers feel it first. This visual explains why. Through my experience, I’ve learned a hard truth: Throughput is not the sum of efficiencies—it is controlled by one constraint. What this bottleneck analysis really shows 1️⃣ Capacity Upstream ≠ Throughput Downstream You can widen capacity everywhere: - Faster suppliers - Bigger supermarkets - Higher utilization in Process A None of it matters if one step produces slower than takt. The hourglass doesn’t lie. 2️⃣ Takt Time Is the Customer’s Voice Takt time is not an internal target. It’s the market pulling on your system. When any process: Has capacity < takt Suffers instability or downtime It becomes the constraint—whether you label it or not. 3️⃣ The Bottleneck Is the Revenue Gate Every minute lost at the bottleneck is: - Lost throughput - Lost sales - Lost trust WIP piles up before it. Starvation happens after it. And leaders often chase symptoms in both directions. 4️⃣ Local Optimization Makes the Constraint Worse Speeding up non-bottlenecks: - Increases inventory - Hides the real problem - Creates false confidence The system doesn’t need more effort. It needs constraint focus. 5️⃣ Flow Stops Where Discipline Stops Downtime, stoppages, queues, and withdrawals don’t happen randomly. They happen when: - Capacity planning ignores variability - Flow decisions aren’t constraint-led Management attention is spread evenly instead of intentionally Why this matters High-performing plants don’t ask: “How do we improve everything?” They ask: “What limits us right now—and how do we protect it?” Because when the constraint flows: - Lead time collapses - WIP stabilizes - Revenue follows The rest of the system naturally falls into line. The best operations don’t chase utilization. They design flow around the constraint. If this resonates, happy to exchange notes on real-world impact and ROI. Curious question to leave you with: In most plants, the bottleneck is known—but not addressed. Is that what you see as well?

When Industry 4.0 meets your mud engineer, expect a leap in efficiency and insight. With real-time data, machine learning, and automation, routine tasks like mud checks are transformed into continuous monitoring, allowing engineers to focus on strategic decisions rather than manual checks. Collaborative intelligence combines their expertise with AI’s power, creating a digital twin of your mud engineer’s workflow. This enables proactive management of solids removal, dilution economics, and waste—driven by real-time diagnostics and predictive analytics. Engineers gain deeper visibility into well conditions, optimizing drilling performance and reducing Non-Productive Time (NPT). Industry 4.0 empowers mud engineers with the tools needed to make informed decisions swiftly, setting a new standard in digital fluids management that enhances operational performance and well integrity.