Why LinePlan Rebuilt Its MES Screens Around ISA-101 After a SaaS UI Flopped

The breaking point came during a night shift at a steel meltshop and hot rolling mill project. We had SAP S/4HANA on the enterprise side and Siemens S7 PLCs plus Level-2 automation servers on the shop floor. Yet the production planner still had 4 spreadsheets open at 2:00 AM trying to figure out which heats in the 1640°C reheating furnace matched the next rolling campaign. A single wrong billet charged into the furnace meant 18 tons of downgraded steel and a 45-minute cobble delay. I saw the exact same disconnect in a 12-line FMCG hygiene factory in Belgium and Turkey, where operators had to click through 9 ERP screens just to confirm a pallet or split a 12-minute downtime event. Existing tools were either rigid enterprise modules that took 14 months to customize or generic SaaS dashboards that knew nothing about sequence-dependent setups, FIFO reel aging, or OPC UA tag debouncing. I built LinePlan (https://lineplanai.com) to put finite capacity APS scheduling, ISA-101 operator PODs, 3D WebGL factory digital twins, and 5,280-bin 3D warehouse maps on one unified data model.

What worked best for winning paying plant subscriptions was stripping away every Book a Demo gate, putting 5 full-scale interactive simulations live in the browser at https://lineplanai.com/en/demo, and publishing transparent per-plant recurring SaaS pricing ($1,950 to $8,500 per month per plant with unlimited viewer seats) at https://lineplanai.com/en/pricing. Any production engineer, shift supervisor, or plant director can immediately open our 16-screen MES and SCADA cockpit, our 3D Iron and Steel Meltshop Digital Twin, our 12-line FMCG packaging twin, or our 5,280-bin 3D Warehouse map without typing an email address, and then move straight into a 14-day guided plant sandbox using their own BOM, work center, and routing data. Our telemetry shows engineers spend an average of 7 minutes and 1 second inside the MES simulator before starting a plant sandbox. Publishing deep technical architecture guides in English, German, and Turkish also drove high-intent organic search traffic across 73 industrial queries. What completely flopped early on was generic startup landing page copy and per-seat pricing. Factories hate per-seat fees on the shop floor, so switching to flat per-plant monthly and annual SaaS billing immediately unlocked conversions.

Our industrial users complain about two things above everything else, shop-floor screen latency during shift changes and ERP inventory mismatches when a PLC or network link drops mid-shift. On a 1366x768 touch panel next to a furnace or converting line, an operator wearing protective gloves cannot tolerate a web screen that takes 4 seconds to load or requires scrolling through 20 columns. We solved the ergonomics issue by redesigning all 16 MES screens around strict ISA-101 high-contrast HMI standards, where every core action, scanning a FIFO component batch, splitting a PLC downtime event, or printing a pallet SSCC label, takes at most 3 taps with zero page scroll. To solve the mid-shift network drop complaint, we built an idempotent store-and-forward queue between the edge OPC UA / SAP PCo layer and ERP. Even if the WAN or ERP goes offline for 2 hours during maintenance, local operator PODs keep running at 60 FPS and replay every 101 goods receipt and 261 component consumption confirmation in exact timestamp order once connectivity returns.

Most software builders think production planning and MES is about drawing a colorful Gantt chart or collecting OEE percentages on a TV screen. That is barely 15% of the real problem. In a real factory, a schedule breaks 20 minutes into the shift because a spectrometer test fails on Heat 402, a slitter blade wears out early, or 40 pallets in aisle 4 are put on quality hold. If your APS scheduler does not know the real-time quarantine status of bins in the warehouse, or if your MES cannot enforce a physical poka-yoke interlock when an operator scans a non-FIFO reel, the software becomes shelfware within 30 days. Another huge misconception is that factories want "AI magic" to auto-pilot their plant. Production planners in steel, paper, and automotive plants never trust a black box. They want deterministic constraint solvers that show the exact setup penalty in minutes, highlight which 3 downstream orders will slip if a rush order is inserted, and let the human planner approve the final schedule diff.

Over the next 6 to 12 months, we are rolling out three major capabilities on https://lineplanai.com. First, we are expanding our pre-packaged integration adapters for SAP Digital Manufacturing (SAP DM / DMC), SAP MII, and SAP S/4HANA (PP, MM, PM, QM, and EWM) so multi-plant manufacturers can deploy our 3D Digital Twin and 3D Warehouse layers alongside their existing ERP in under 4 weeks. Second, we are adding a self-service 3D warehouse layout generator that converts standard ERP bin master tables (Aisle, Rack, Level, Bin) into an interactive 60-FPS WebGL high-bay warehouse with live FIFO aging heatmaps in 1 day. Third, we are releasing a multi-constraint campaign optimizer for steel meltshops and paper converting lines.

We currently have 6 paying manufacturing plants on recurring monthly and annual LinePlan SaaS subscriptions across steel, paper converting, and discrete assembly, generating $21,400 in Monthly Recurring Revenue (MRR) at an average subscription of $3,566/month per plant (our recurring tiers run from $1,950 to $8,500/month per plant at https://lineplanai.com/en/pricing). Across those 6 subscribed plants, 48 production lines and 310 active shop-floor operator and planner users run LinePlan every shift, with 100% net retention since launch.

Before building LinePlan, I spent more than 10 years working hands-on on the factory floor as an SAP MES Solution Architect and Digital Manufacturing Lead. Over that decade, I delivered 11 full-cycle MES implementations (5 SAP MII, 3 SAP ME, and 3 SAP DM / DMC projects) plus 5 live plant support engagements across 16 manufacturing plants in 8 countries, including Germany, Belgium, Italy, the Netherlands, Romania, Greece, Algeria, Morocco, and Turkey. I wrote the actual MII BLS transactions, HANA SQLScript procedures, SAP UI5 operator PODs, and OPC UA / Level-2 integrations for electric arc furnaces, continuous casting machines, hot rolling mills, tissue converting lines, and white-goods assembly plants. LinePlan was born directly out of those 23+ custom shop-floor modules built in the field.

My biggest mistake early on was designing the first version of our web interface like a typical modern B2B SaaS app, lots of whitespace, rounded cards, and decorative charts. The first time I put it in front of a veteran production planner and a shift supervisor, they hated it. On a control room monitor, whitespace means you can only see 6 work orders instead of 35, and soft pastel colors wash out under factory fluorescent lighting. We threw out the SaaS template and rebuilt both the APS workbench and the 16-screen MES cockpit from scratch using ISA-101 industrial HMI principles, dark slate and high-contrast tabular grids, monospace numeric alignment, explicit state badges, and instant keyboard and touch shortcuts. Engagement time immediately jumped past 6 minutes per session.