A plant manager's guide to Manufacturing Execution Systems and Enterprise Resource Planning — when ERP alone is enough, when MES fills the gap, and when you need both integrated across the ISA-95 stack. Built from 20+ years of implementing shop floor and business systems for discrete and process manufacturers across the Midwest.
Here is what happens at a 150-person discrete manufacturer running Epicor Kinetic as their ERP with no MES layer. A customer calls at 10:15 AM asking for a status update on their order. The customer service rep opens Epicor, finds the sales order, sees that the linked work order was released to the floor three days ago. The status says 'In Progress.' That is all the information the ERP has. To answer the actual question — which operation is the job on right now, how many pieces are complete, and will it ship on time — the CSR walks to the production office, where the scheduler pulls up an Excel tracker, calls the floor supervisor on a two-way radio, and waits while the supervisor physically locates the job traveler. Fifteen minutes later, the answer comes back: the job is on its second operation, 60% through the run, and it will probably make the ship date unless the CNC that is currently running it throws another alarm. The CSR relays this to the customer and logs it nowhere. The next time someone asks about the same order, the entire process repeats.
This is not a technology failure. It is an architecture gap. ERP systems — SAP S/4HANA, Microsoft Dynamics 365 Finance and Operations, Epicor Kinetic, Infor CloudSuite Industrial, Sage X3 — are designed for ISA-95 Level 4: business planning and logistics. They manage financial transactions, purchase orders, sales orders, material planning (MRP), and inventory valuation. They work in transactional time: orders are entered, MRP runs nightly or weekly, work orders are released in batches, and job costs are calculated after the fact. ERP systems are excellent at answering 'What should we build, when should we build it, and what did it cost?' They are structurally incapable of answering 'What is happening on the shop floor right now, this second?'
Manufacturing Execution Systems occupy ISA-95 Levels 2 and 3 — the layers between shop floor automation (PLCs, SCADA, sensors) and business planning (ERP). MES systems like Plex (now Rockwell Automation's cloud MES), DELMIAworks (formerly IQMS), Aegis FactoryLogix, Rockwell FactoryTalk ProductionCentre, GE Proficy, and MPDV Hydra handle what happens between 'work order released' and 'finished goods received into inventory.' They track real-time machine status, operator clock-in/clock-out per operation, piece counts by shift and work center, scrap and rework with reason codes, SPC data collection for in-process quality, and actual cycle times versus standard times. Without MES, your ERP has a blind spot that covers 100% of what happens during production — the most expensive and variable part of your entire operation.
The financial impact of this blind spot is measurable. A 2023 MESA International study found that manufacturers without MES-level shop floor data collection understate actual production costs by 12-18% because labor and scrap are estimated rather than captured in real time. A plant running 50 jobs per week with an average job value of $8,000 and a 15% cost understatement is misquoting $60,000 in margin per week — $3.1 million per year. That is not theoretical waste. It is money walking out the door on every quote because your job costing is based on standards that were set in a routing table two years ago and never validated against actual floor performance.
ERP shows work orders 'In Progress' with zero visibility into which operation, machine, or operator is running the job right now
Job costing based on routed standards, not actual labor and cycle times — understating true production cost by 12-18%
No real-time OEE data: machine utilization, performance, and quality calculated manually from shift reports days after the fact
Scrap and rework tracked on paper travelers or not tracked at all — root cause analysis is impossible without timestamped defect data
Production scheduling in Excel because ERP finite scheduling requires real-time capacity data that does not exist without MES
Customer order status requires phone calls to the floor supervisor instead of a dashboard query — 15-minute answer to a 5-second question
SPC and quality data collected on paper checksheets, then manually entered into a quality system that is disconnected from production records
Our engineers have built this exact solution for other businesses. Let's discuss your requirements.
The ISA-95 standard (ANSI/ISA-95, also IEC 62264 internationally) defines the integration architecture between enterprise systems and manufacturing operations. Understanding this model is the single most important framework for deciding whether you need MES, ERP, or both. The model has five levels. Level 0 is the physical production process — metal being cut, plastic being injected, chemicals being mixed. Level 1 is sensing and manipulation — PLCs, sensors, motor drives, and actuators that control the physical process. Level 2 is monitoring and supervisory control — SCADA systems, HMIs, and DCS platforms that operators use to monitor and intervene in the process. Level 3 is Manufacturing Operations Management — this is the MES layer. It handles work order dispatching, detailed production scheduling, real-time data collection, quality management, performance analysis, and labor tracking. Level 4 is Business Planning and Logistics — the ERP layer. It handles financial management, supply chain planning, customer order management, and enterprise-wide resource planning.
The critical boundary is between Level 3 and Level 4. ERP (Level 4) decides what to produce and when, based on demand forecasts, customer orders, material availability, and capacity constraints modeled at a rough-cut level. MES (Level 3) decides how to produce it — which work center runs which operation, in what sequence, with which operator, using which tooling, at what speed, and with what quality checkpoints. ERP releases work orders. MES dispatches them. ERP plans material requirements. MES confirms actual material consumption. ERP estimates job costs from standard routings. MES captures actual costs from real-time labor, machine time, and scrap data. The two systems serve fundamentally different time horizons: ERP operates in days, weeks, and months. MES operates in seconds, minutes, and hours.
FreedomDev has been implementing both sides of this boundary for over 20 years. We have deployed ERP systems (Epicor Kinetic, Infor CloudSuite Industrial, Dynamics 365, SAP Business One), built custom MES solutions for plants that needed real-time shop floor tracking without the $500K-$2M price tag of a full commercial MES platform, and most importantly, built the Level 3-to-Level 4 integration layer that connects MES data to ERP business logic. That integration layer is where most MES and ERP projects fail — not because the individual systems do not work, but because the data mapping between them was never properly architected. Work order status, material consumption, labor transactions, scrap reporting, quality holds, and finished goods receipts all need to flow bidirectionally between MES and ERP in near-real-time. When they do not, you end up with two systems that both have data, and neither has the truth.
The core function of any MES system is capturing what actually happens during production — not what the routing says should happen. This means operator login/logout per operation with timestamp accuracy to the minute, piece count tracking (good parts, scrap, rework) by work center and shift, actual cycle time per part versus standard cycle time, machine state tracking (running, idle, setup, unplanned downtime, planned maintenance), and material consumption by lot or serial number. FreedomDev builds shop floor data collection using barcode scanning at each work center, tablet-based operator interfaces, and where the budget supports it, direct machine integration via OPC-UA or MTConnect protocols that pull cycle counts and machine states directly from CNC controllers and PLCs without operator input.
Overall Equipment Effectiveness is the standard manufacturing KPI, calculated as Availability times Performance times Quality. Availability is planned production time minus unplanned downtime, divided by planned production time. Performance is actual cycle time divided by ideal cycle time. Quality is good parts divided by total parts produced. A world-class OEE is 85%. Most manufacturers without MES-level data collection do not know their real OEE — they estimate it from shift reports completed at the end of the day, which consistently overstate actual performance by 10-20%. FreedomDev builds real-time OEE dashboards that calculate availability, performance, and quality per machine, per work center, and plant-wide, updated every production cycle. These dashboards mount on large screens on the shop floor so operators and supervisors see the numbers in real time, not in a report tomorrow morning.
ERP-level MRP generates a list of work orders that need to run. MES dispatches them — assigns each operation to a specific work center, in a specific sequence, with priority rules that account for due date, customer priority, setup time minimization (grouping similar jobs to reduce changeovers), tooling availability, operator certification requirements, and real-time machine status. Without MES dispatching, schedulers build the dispatch list manually in Excel or on a whiteboard, and any disruption — a machine down, a rush order, a material shortage — requires manual replanning that takes hours. MES-level dispatching reruns the schedule in minutes using constraint-based algorithms and presents the revised plan to the floor supervisor for approval.
Statistical Process Control is the practice of monitoring dimensional and attribute data during production to detect drift before it produces out-of-spec parts. Traditional SPC uses paper checksheets — an operator measures every 10th part with a caliper, writes the measurement on a form, and a quality engineer plots it on a control chart sometime next week. MES-integrated SPC captures measurements electronically at the point of inspection — either via manual gage input on a tablet, digital gage output fed directly into the system, or CMM (coordinate measuring machine) integration. Measurements are plotted on X-bar/R charts, p-charts, or c-charts in real time. When a measurement exceeds control limits, the system triggers an immediate alert to the quality engineer and can optionally halt the machine. This catches process drift within minutes instead of days, reducing scrap from out-of-control processes by 30-60%.
The integration between MES and ERP is the most technically challenging and most frequently botched part of a manufacturing systems architecture. Data flows in both directions. From ERP to MES: released work orders with bill of material, routing, and customer due date; inventory allocations and material availability; engineering change orders that modify BOMs or routings mid-production. From MES to ERP: actual labor hours per operation for job costing; scrap and rework quantities with reason codes; material consumption (actual versus planned) for inventory accuracy; machine hours for overhead absorption; quality hold status for lot dispositioning; and finished goods receipt to close the work order and update inventory. FreedomDev builds this integration using RESTful APIs, message queues, or direct database integration depending on the ERP platform. We map every data element, define the sync frequency (real-time, near-real-time, or batch), and build reconciliation logic to handle conflicts when the same record is modified in both systems.
Regulated industries — aerospace (AS9100), automotive (IATF 16949), medical devices (FDA 21 CFR Part 11), food and beverage (FSMA) — require full traceability from raw material receipt through finished goods shipment. This means tracking which lot of raw material went into which work order, which operator ran which operation, on which machine, at what time, with what inspection results. ERP systems track lot and serial numbers at the inventory transaction level. MES systems track them at the operation level — capturing the full genealogy of every component and process step that contributed to a finished unit. When a customer reports a defect, a proper MES traceability system can identify every unit affected by the same material lot, the same machine, or the same process condition in minutes rather than days.
We had Epicor for six years and never turned on shop floor control because our last implementer said it was too complex. FreedomDev enabled it, connected it to barcode scanning at every work center, and built a real-time OEE dashboard in 10 weeks. For the first time in the history of this plant, I can tell a customer exactly where their order is without picking up a radio.
We map your current systems architecture against the ISA-95 model. What lives at Level 4 (ERP, financials, supply chain)? What lives at Level 3 (shop floor tracking, scheduling, quality)? What is missing entirely? For most mid-market manufacturers, the assessment reveals a solid ERP layer with a complete void at Level 3 — production data lives in Excel, paper travelers, and whiteboards. We document every data flow between the shop floor and the business systems, identify the gaps, and quantify the cost of those gaps in labor, scrap, misquoted jobs, and late deliveries. Deliverable: a gap analysis with a prioritized roadmap that tells you exactly whether you need full MES, targeted MES modules, ERP enhancements, or custom middleware.
Based on the gap analysis, we present the options with hard cost comparisons. Option A: commercial MES platform (Plex, DELMIAworks, Aegis) with ERP integration — typically $300K-$2M+ all-in for a 100-200 person plant, 6-18 month implementation. Option B: ERP enhancement using your existing platform's manufacturing modules more aggressively — enabling shop floor reporting in Epicor, activating MES features in Infor SyteLine, or deploying Dynamics 365 Supply Chain Management's production floor execution module — typically $50K-$200K and 2-6 months. Option C: custom MES or targeted MES modules (shop floor data collection, OEE dashboards, dispatch boards) built to integrate with your existing ERP — typically $80K-$400K depending on scope, 3-8 months. We recommend based on your plant size, complexity, regulatory requirements, and budget. There is no single right answer.
Before any build begins, we design the data architecture that connects MES-level and ERP-level systems. This means defining every data element that crosses the Level 3 / Level 4 boundary: work order release, operation start/stop, piece counts, scrap codes, labor transactions, material issues, quality results, and finished goods receipt. We specify the sync method for each element (API call, message queue, database trigger, or batch file), the frequency (real-time, every N minutes, or daily batch), and the conflict resolution rules for bidirectional data flows. We also define the master data governance — which system owns the item master, which owns the routing, which owns the work center definitions — because master data conflicts are the number one cause of MES-ERP integration failures.
Implementation runs in parallel tracks: MES configuration or custom development, ERP configuration changes, integration middleware development, shop floor hardware setup (barcode scanners, tablets, display screens), and user training. We deploy to one production cell or work center first as a pilot — not a conference room pilot with test data, but a live production cell running real jobs. Operators, supervisors, and schedulers use the system during actual production while we stand on the floor and fix issues in real time. The pilot cell runs for 2-4 weeks. Every defect, every workflow friction point, and every 'this is not how we actually do it' conversation gets resolved before we expand to the full plant.
After the pilot cell validates the system, we roll out work center by work center across the plant. Each rollout phase includes operator training (role-based, at the work center, during the shift — not in a conference room), supervisor training on the dispatch board and exception handling, and management training on the OEE dashboards, job costing reports, and traceability queries. Post-rollout, we provide 60 days of on-site and remote support. After stabilization, we transition to a continuous improvement engagement: monthly reviews of OEE trends, scrap Pareto analysis, scheduling efficiency, and identification of the next bottleneck to target. Manufacturing operations software is never 'done' — the plant changes, products change, and the system must evolve with them.
| Metric | With FreedomDev | Without |
|---|---|---|
| Decision Scope | MES vs ERP vs both — we assess and recommend the right architecture for your plant, not a product | MES vendors sell MES. ERP vendors sell ERP. Neither assesses whether you actually need their product. |
| Integration Depth | Bidirectional Level 3-to-Level 4 integration designed before implementation begins | Integration treated as an afterthought — bolted on after both systems are deployed separately |
| Shop Floor Reality | Pilot cell on live production jobs with real operators before plant-wide rollout | Conference room pilot with test data — first real-world use is go-live day |
| Commercial MES (Plex, DELMIAworks) | We implement, configure, and integrate commercial MES platforms with your existing ERP | MES vendor implements their system; a different firm handles ERP integration (finger-pointing when it fails) |
| Custom MES Modules | Build only the MES modules you need (OEE, dispatch, SPC) — $80K-$400K | Full commercial MES platform whether you need every module or not — $300K-$2M+ |
| ERP-Only Enhancement | Maximize your existing ERP's manufacturing modules before adding MES — lowest cost path | ERP vendor's professional services team lacks shop floor implementation experience |
| Master Data Governance | Defined in architecture phase: which system owns each master record, with conflict resolution rules | Discovered as a problem 6 months into implementation when data conflicts start corrupting both systems |
| Ongoing Optimization | Monthly OEE analysis, scrap Pareto, scheduling efficiency reviews, and continuous improvement roadmap | Go-live is the end of engagement — optimization is your problem |
Schedule a direct technical consultation with our senior architects.
Make your software work for you. Let's build a sensible solution.