https://amit-personal-site.pages.dev/ Insights on Data, AI, and Digital Transformation for Manufacturing, FMCG, Packaging, and Logistics operations — by Amit Kumar Singh, MyData Insights. en-gb © 2026 MyData Insights Pvt Ltd amit@mydatainsightspvtltd.com (Amit Kumar Singh) amit@mydatainsightspvtltd.com (Amit Kumar Singh) 2026-05-12T00:00:00+00:00 https://amit-personal-site.pages.dev/profile.jpg https://amit-personal-site.pages.dev/ https://amit-personal-site.pages.dev/blog/blog-1 https://amit-personal-site.pages.dev/blog/blog-1 2026-05-08 Most ERP reports show you last night\ amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-2 https://amit-personal-site.pages.dev/blog/blog-2 2026-04-15 Unplanned downtime costs US manufacturers $50 billion a year. Predictive maintenance can cut that — but the gap between vendor claims and shop-floor reality is significant. Here\ amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-4 https://amit-personal-site.pages.dev/blog/blog-4 2026-04-01 European manufacturers leave €50B+ on the table annually — not from lack of effort, but from lack of analytical visibility into what drives OEE, quality, and throughput. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-3 https://amit-personal-site.pages.dev/blog/blog-3 2026-03-22 FMCG companies spend millions on AI-powered demand planning tools and still run 8% stockout rates. The AI is not the problem. The data underneath it is. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-5 https://amit-personal-site.pages.dev/blog/blog-5 2026-03-10 91% of CIOs cite data silos as a top obstacle to digital transformation. The root cause: treating the ERP as an analytics platform when it was designed as a transaction system. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-6 https://amit-personal-site.pages.dev/blog/blog-6 2026-03-01 68% of IIoT projects stall due to OT/IT integration failures. Not because the technology does not work — because the architecture decisions and organisational model were wrong from the start. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-7 https://amit-personal-site.pages.dev/blog/blog-7 2026-02-22 The average FMCG company runs 7–12 disconnected systems. The cost shows up as stockouts, overstock, missed promotions, and S&OP meetings that debate whose numbers are right. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-8 https://amit-personal-site.pages.dev/blog/blog-8 2026-02-15 Microsoft Fabric consolidates data engineering, real-time analytics, and Power BI into a single platform. Here is an honest assessment of where it delivers — and where it does not — for manufacturing operations. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-9 https://amit-personal-site.pages.dev/blog/blog-9 2026-02-05 Poor data quality costs organisations $12.9M per year on average. In manufacturing, the problem is worse — because OT data has no natural owner, KPIs are contested, and context is implicit. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-10 https://amit-personal-site.pages.dev/blog/blog-10 2026-01-28 Vendor PdM demonstrations make it look straightforward. The labelled training data problem, the CMMS integration gap, and the asset selection mistake are what actually determine success or failure. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-11 https://amit-personal-site.pages.dev/blog/blog-11 2026-01-20 82% of manufacturers report OEE figures that undercount actual losses. The miscalculation is not random — it systematically misallocates improvement investment and leaves real throughput gain uncaptured. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-12 https://amit-personal-site.pages.dev/blog/blog-12 2026-01-12 AI forecasting tools can reduce forecast error by 50% — but only on a clean, unified demand signal. On bad data, they produce confident wrong answers that drive worse supply chain decisions than a spreadsheet. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-13 https://amit-personal-site.pages.dev/blog/blog-13 2026-01-05 Quality failures cost manufacturers 15–20% of revenue annually. Pass/fail reporting tells you what happened. Root cause analytics — connecting process data, material attributes, and quality outcomes — tells you why, and how to prevent it. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-14 https://amit-personal-site.pages.dev/blog/blog-14 2025-12-28 Most organisations have a visibility dashboard and call it a control tower. There are two more layers of value above visibility — intelligence and automation — and the gap between them represents 35% fewer supply chain disruptions. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-15 https://amit-personal-site.pages.dev/blog/blog-15 2025-12-20 The gap between data-driven factories and the rest is widening. Closed-loop operations — where defined signals automatically trigger defined responses — are how leading manufacturers are pulling ahead. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-16 https://amit-personal-site.pages.dev/blog/blog-16 2025-12-12 $984 billion in excess inventory sits in consumer goods supply chains globally. Meanwhile, stockout rates average 8%. Both problems have the same root cause — and automated replenishment addresses both. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-17 https://amit-personal-site.pages.dev/blog/blog-17 2025-12-04 RPA delivers 200–400% ROI when applied to the right processes — and near-zero ROI when applied to the wrong ones. The difference between the two is specific, identifiable, and often overlooked in RPA planning. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-18 https://amit-personal-site.pages.dev/blog/blog-18 2025-11-25 65% of logistics SLA breaches are detectable in the data more than 4 hours before the customer is impacted. The technology to fix this is not complicated. Here is the architecture that makes it work. amit@mydatainsightspvtltd.com (Amit Kumar Singh) https://amit-personal-site.pages.dev/blog/blog-19 https://amit-personal-site.pages.dev/blog/blog-19 2025-11-18 87% of AI and analytics projects never reach production operational impact. The gap is almost never the model quality. It is the missing bridge between insight on a dashboard and action in an operational workflow. amit@mydatainsightspvtltd.com (Amit Kumar Singh)