Manufacturing Execution Systems (MES) and ERP platforms are the operational backbone of modern manufacturing. Migrating them to the cloud is technically demanding, regulatorily complex and operationally risky — production cannot stop. This playbook describes the proven migration approach for manufacturing IT on AWS: from structured discovery through dependency mapping to meticulously planned cutover. Special focus is given to FDA 21 CFR Part 11 compliance for pharmaceutical and medical device companies and to maximizing AWS MAP funding.

Why MES/ERP Migrations in Manufacturing Are Uniquely Complex

An ERP migration at a retail company is complex. An ERP migration at a manufacturing company is a different magnitude. Manufacturing IT is not an island: MES and ERP are deeply integrated with OT systems (PLCs, SCADA, quality control), supplier portals, logistics platforms and reporting infrastructure. A single missed interface can turn the cutover into a disaster.

Added to this is time pressure: manufacturing companies often have narrow maintenance windows for system changeovers — weekends, plant shutdowns or planned line stops. If the migration exceeds that window, production runs without critical systems. In most cases that is not tolerable.

Key Terms

Manufacturing Execution System (MES)
Software that controls, monitors and documents production processes in real time — from order release to completion confirmation. The MES mediates between the planning system (ERP) and the production floor (OT). Typical functions: order management, machine data acquisition, quality assurance, traceability.
Dependency Mapping
Systematic capture of all technical dependencies between applications, databases, middleware components, interfaces and infrastructure. Dependency mapping is the foundation for safe migration planning and prevents surprising outages after cutover.
Cutover
The planned switch from the old to the new system — the moment production first accesses the cloud environment. The cutover plan contains a minute-by-minute step-by-step guide, go/no-go criteria and a rollback plan.
FDA 21 CFR Part 11
US regulation for electronic records and electronic signatures in the pharmaceutical industry. Requirements include tamper-evident audit logs, access controls, electronic signatures with authorization proof and system validation (IQ/OQ/PQ).

Phase 1: Discovery and Dependency Mapping

No migration without a complete inventory. The first step is a systematic discovery of all applications, infrastructure components and interfaces. AWS provides specific tooling:

  • AWS Application Discovery Service: Installs a lightweight agent on servers to automatically capture running processes, network connections, resource usage and configuration data.
  • AWS Migration Evaluator: Provides a detailed analysis of the existing infrastructure and generates an optimized cost model for AWS.
  • Dependency workshops: Facilitated sessions with application owners to validate automatically discovered dependencies and capture undocumented interfaces.

Frequently Overlooked Dependencies in Manufacturing

System Category Examples Migration Risk
OT Systems SCADA, PLC gateways, quality inspection systems High — often proprietary protocols, no API documentation
Database servers Oracle, MS SQL, SAP HANA Medium — licensing questions, replication strategies
Middleware / ESB SAP PI/PO, MuleSoft, IBM MQ High — central integration hub with many dependent systems
Reporting / BI SAP BW, Power BI, QlikSense Medium — often live database connections rather than APIs
Authentication Active Directory, LDAP, SAP IdM High — single point of failure for all systems

Phase 2: Choosing a Migration Strategy (The 7 Rs)

  1. Retire: Decommission the application — no longer providing value.
  2. Retain: Keep on-premises — systems with hard real-time requirements.
  3. Rehost (Lift & Shift): Move servers 1:1 as EC2 instances. Fastest method, lowest risk.
  4. Replatform (Lift, Tinker & Shift): Minor optimizations during migration — e.g. Oracle database to Amazon RDS for managed patching.
  5. Repurchase: Switch to SaaS — e.g. replace on-premises MES with a cloud-native MES.
  6. Refactor / Re-architect: Fundamentally modernize the application — highest effort, greatest cloud benefit.
  7. Relocate: Move VMware workloads 1:1 to VMware Cloud on AWS.

For most manufacturers, Rehost + Replatform is the preferred strategy in the first migration wave. Critical MES components with hard latency requirements are initially retained on-premises or operated on AWS Outposts.

Phase 3: The Cutover Plan — Minute by Minute

  1. Define cutover window: When is the best time from a production standpoint? Typically Friday evening after shift end, start of plant shutdown, between two production campaigns.
  2. Pre-cutover checklist (D-7 to D-1): Verify all database replications. Final acceptance testing of the AWS environment. Practice the rollback procedure. Activate the communication plan.
  3. Cutover sequence (T-0): Freeze production systems in defined layers (write lock). Final data synchronization MES database → AWS. Switch DNS records. Start MES/ERP on new infrastructure. Run integration tests in the live system. Go/No-Go decision by steering committee.
  4. Hypercare phase (D+1 to D+14): Elevated support readiness. Daily status meetings. Proactive CloudWatch monitoring. Fast escalation paths.
  5. Rollback plan (if No-Go): Clear criteria for when rollback is triggered. Maximum time buffer defined (e.g. T+4 hours = latest rollback decision point).

FDA 21 CFR Part 11: Cloud Compliance for Pharma and MedTech

AWS provides a strong foundation for FDA 21 CFR Part 11 compliance:

  • Tamper-evident audit logs (AWS CloudTrail): All API calls are logged and can be stored immutably in S3 with Object Lock (WORM) — exactly what 21 CFR Part 11 requires for electronic records.
  • Access controls (AWS IAM): Granular permission management per policy. Every user authenticates with a unique identity. MFA mandatory for privileged accounts.
  • Electronic signatures (AWS KMS): Cryptographically secure electronic signatures linked to CloudTrail.
  • System validation (IQ/OQ/PQ): AWS provides documentation for Installation Qualification (IQ). Storm Reply supports the complete validation process.

Storm Reply: Migration Expertise for Critical Manufacturing Systems

Storm Reply, as an AWS Premier Consulting Partner with AWS Migration Competency, offers: Migration Readiness Assessment, Dependency Mapping workshops, minute-by-minute cutover plans with rollback scenarios, validation support (IQ/OQ/PQ) for regulated industries and MAP funding application support.

Frequently Asked Questions

What is an MES and why is cloud migration complex?
A Manufacturing Execution System controls and monitors production processes in real time. MES systems often have hard latency requirements, many interfaces to OT systems and are deeply integrated into production processes — making a phased approach necessary.
How long does an ERP cloud migration in manufacturing take?
A lift-and-shift migration typically takes 3–6 months. A comprehensive modernization can take 12–24 months. Planning the cutover phase is often the most time-critical part.
How does AWS support FDA 21 CFR Part 11 compliance?
AWS provides CloudTrail for tamper-evident audit logs, AWS KMS for electronic signatures and IAM for granular access controls. Storm Reply supports the validation process (IQ/OQ/PQ) for pharmaceutical and medical device customers.
What is dependency mapping and why is it so important?
Dependency mapping captures all technical dependencies between applications, databases, interfaces and infrastructure. Missing dependencies are the most common cause of failed cutovers. AWS Application Discovery Service automates this process.

Migration Readiness Assessment

How ready are your MES and ERP systems for the cloud? Let us conduct a structured assessment together — including MAP funding review.

Get in touch

More Insights