mtabusa - Blog #Smart Manufacturing

How SIRI and the Manufacturing Technology Balance Sheet Guide Technology Sequencing

Tue, 06 Jan 2026 17:00:00 +0000

Investing in Smart Manufacturing with Discipline

Most manufacturers have no shortage of technology options. They are short of time, attention, and capacity to absorb change. The real question is not “What should we buy next” but:

How do we sequence smart manufacturing investments so that we reduce chaos, avoid change fatigue, and move our operational and financial metrics in the right direction?

For CEOs, COOs, and owners, the way you sequence smart manufacturing investments will shape three decisions:

Which projects to fund now versus postpone
Where to say “not yet” to attractive technologies that are ahead of your readiness
How to allocate scarce leadership attention and capital over the next 12 to 18 months

I build on earlier blogs where I wrote about the Manufacturing Technology Balance Sheet (MTBS), digital twins as decision tools, and a practical roadmap for small manufacturers.

Sequencing is critical to avoid bottlenecks and change fatigue

Poor sequencing shows up in very familiar ways:

The same process engineers and supervisors are pulled into three initiatives at once
New systems land on top of unstable processes, so people work around them
Everyone is “in workshops,” but OEE, lead times, and on time delivery remain unchanged
Operators see new tools as overhead.

Smart manufacturing initiatives create load along three dimensions:

Process load – changes in work instructions, routings, approvals, and responsibilities
Data load – new fields, new codes, new data entry expectations, and new system behaviors
People load – training, pilots, debugging, meetings, and mental overhead

If you stack high load projects on the same people and the same processes at the same time, you create real risk.

Sequencing is risk management. When you see two or more high load projects landing on the same roles or processes in the same quarter, it is a signal to reorder or resize the portfolio before resistance shows up as a performance problem.

Linking sequencing to the Manufacturing Technology Balance Sheet

The MTBS lens keeps you honest. Every initiative should be traceable to one or more of these questions:

Where are we losing capacity on critical assets
Where are we carrying unnecessary inventory or WIP
Where are we seeing rework, warranty, or scrap that erodes margin
Where are we exposed to workforce risk (retirements, skill gaps, training load)

A simple way to link sequencing to MTBS:

Rank value at stake.
For each area (capacity, quality, working capital, workforce), ask your finance lead for an order of magnitude estimate.
Identify enabling dependencies.
Ask your team: “What must be true before this initiative can sustain?” For example, a scheduling optimizer depends on trustworthy routings and cycle time data.
Place initiatives into three buckets.

Foundation: fixes that reduce chaos and unlock future options (for example, basic connectivity, event logging, BOM and routing clean up)
Multiplier: initiatives that amplify existing strengths (for example, expanding successful OEE pilots, digital work instructions in a stable line)
Optional: ideas that are attractive, but not yet supported by process or data maturity

Examples of phased implementation

Example 1: CNC Machine Builder OEM
Problem: Complex engineered to order machines, long lead times, frequent engineering changes, and heavy dependence on a few experts.
Phased sequence: Simplify design CAD structure into modules and emphasize reuse; Use simulation twin reference model to de-risk automation module and shorten commissioning; Connect field performance and service data for XaaS that impact you and your customer.

Outcome: At MTAB Engineers, our cost to serve was reduced by 10%.

Example 2: High mix component manufacturer
Problem: Many SKUs, frequent changeovers, unstable schedules, and constant firefighting by planners and supervisors.
Phased sequence: Instrument one bottleneck area and standardize changeover routines; Implement OEE and schedule reviews for that area and test batch sizes with a performance twin; Extend successful patterns to adjacent lines and link into planning.

Outcome: Typically. production capacity expanded by 15+%, without additional capital or manpower spend.

How I use SIRI and MTBS to structure a sequencing engagement

For clients, I combine SIRI and the Manufacturing Technology Balance Sheet into a three step service that answers three practical questions.

Where are we now

SIRI maturity assessment for you in relation to your industry
MTBS view of capacity, quality, inventory, and workforce risk

What should we do first

Shortlist of initiatives and sequencing logic based on dependencies and value at stake
Selection of one or two use cases for the next 3 to 9 months

How do we execute without chaos

Clear business and technical owners
Defined scope, success metrics, and a 90 to 120 day Phase 1 plan with explicit entry and exit criteria for Phase 2

This structure turns smart manufacturing into a disciplined capital allocation and capability building exercise.

Start with a SIRI led readiness and sequencing review

If you are considering new investments to expand manufacturing capabilities, the first step is a structured view of your maturity and your value at stake

I offer a SIRI based Smart Manufacturing Readiness and Sequencing Assessment that:

Benchmarks your current state across key Industry 4.0 dimensions in your industry
Links maturity gaps to MTBS metrics such as capacity, quality, and working capital
Prioritizes and sequences your next 12 to 18 months of technology investments
Defines one or two “no chaos” use cases for the next 3 to 6 months with clear success criteria

If you would like to explore a readiness and sequencing review for your operation, you can contact me through MTAB USAor via LinkedIn.

Preparing for Digital Twin Adoption Across Design, Production, and Maintenance

Mon, 08 Dec 2025 05:00:47 +0000

Twin Ready Manufacturing- Preparing Processes, People, and Products

Digital twins are often dismissed as expensive, complex, and hard to sustain. That perception is not about the technology. It is about readiness. When processes are inconsistent, data is unreliable, and ownership is unclear, even a well-built twin becomes a one time experiment instead of a capability that improves decisions.

In my previous blog, I wrote about digital twins as a decision tool and why mid market manufacturers must stay use case driven within a 6 to 12 month window. This post addresses the question I hear most often from manufacturing leaders: How do we become twin ready without turning this into a multi year transformation program?

To make this practical, I will use a real situation from my earlier work. An automotive component manufacturer had an induction hardening bottleneck with manual processes, safety constraints, and prior automation disappointment. We used a digital twin to validate the automation design before deployment, selected a SCARA robot, and improved efficiency by more than 20 percent.

Integration points and what to connect first

Manufacturers stall when they try to connect everything. Anchor integration to one decision, then connect only what improves that decision.

Design and engineering
Connect when the question is: Will this design and process work as expected

CAD geometry and interfaces
Critical characteristics and inspection requirements
Revision history that affects fit, cycle time, or quality

In our project, the part geometry and material informed the decisions for end effector, robot selection and pallet designs.

Manufacturing execution
Connect when the question is: How will this behave under real mix and schedule

Routing and work definitions
Actual cycle times and downtime causes
Exception handling

The model forced clarity on material flow, work steps, and exception handling, which then informed MES logic and operator routines.

Maintenance
Connect when the question is: Can we sustain performance and uptime

Asset hierarchy and critical spares
Failure patterns and mean time to repair
Preventive maintenance compliance for scoped equipment

The automation cell design was influenced by the maintenance team's input on why previous automation failed and what they needed to maintain to high production standards.

Common pitfalls when manufacturers start too early

Building a model before defining the decision it must improve
Expecting the twin to impose process discipline without standard work
Allowing revisions and routing to drift, making the twin obsolete
Over investing in integration before proving one measurable outcome
Chasing high fidelity instead of building credibility and adoption

Induction hardening lesson: The twin worked because it prevented physical rework by resolving constraints before deployment.

Workflow to trial a digital twin (3-6 months)

This workflow is designed to deliver one credible use case and build reusable foundations.

Step 1: Choose one decision

Pick one decision already costing time, margin, or customer confidence.

Examples:

Automation cell design and commissioning risk
NPI ramp stability for the first 8 to 12 weeks
Bottleneck capacity recovery without new equipment
Repeat warranty or service failure on a shipped platform

Step 2: Define the minimum data spine

Create a short list of data required for that decision:

Part and revision identification rule
Operation steps at the level needed for modeling
Cycle time, downtime, and quality outcomes for the scoped area

Step 3: Standardize the local process slice

Standardize only what touches the use case:

Work definition and exception rules
Cause codes that people will actually use
A clear rule for what changes require review

Step 4: Build the twin and validate assumptions

Build the simplest twin that answers the decision:

Simulation twin for layout, reach, interference, cycle time risk
Process twin for NPI ramp or staffing and WIP constraints
Performance twin for a bottleneck schedule and batching rules

Step 5: Operationalize

Weekly review with production, quality, and maintenance
One person accountable for data refresh and model updates
One leader accountable for acting on outputs

Step 6: Expand by reuse

Expand only after the first measurable win:

More variants of the same product family
More assets in the same line
One additional use case that reuses the same data spine

This mirrors your broader transformation principle of starting small, proving value, and scaling based on real impact.

Outcome metrics leaders will recognize and can measure

Choose metrics that both manufacturing and finance teams already track, and that a twin can influence within the scoped use case:

Capacity released on a constrained asset (hours per week or units per shift)
On time delivery for the scoped product family
WIP days or inventory days in the scoped flow

These map to operations reality and working capital impact, and they can usually be measured with existing ERP, MES, and basic production logs.

If you are considering digital twins, start with readiness, not software. I offer a Twin Ready Evaluation that identifies your best first use case, the minimum process and data spine required, and a 90 to 120 day execution plan tied to measurable outcomes. If you want to discuss your first twin use case, write to me.

Related reading:Digital Twins for Competitive Advantage
Related reading:Leveraging Digital Twins for Efficient Automation

From Simulation to Business Model Innovation Using Digital Twins

Wed, 26 Nov 2025 06:29:45 +0000

Digital Twins for Competitive Advantage

How Digital Twins Are Evolving

I think of digital twins in three stages of maturity.

Stage 1: Engineering Simulation Twin

This is where most manufacturers begin.

A virtual model of a machine, cell, or process
Used to validate layout, reach, interference, basic logic, and cycle time
Driven by CAD, offline programming tools, and simulation environments

The value is direct: fewer design rework loops, fewer assembly failures, and fewer surprises during commissioning. This is where many of my own digital twin projects started, especially for automation cells and cost estimation.

Stage 2: Operational Performance Twin

Here, the twin does not retire once the machine is installed.

Connected to runtime data from controls, sensors, quality records, and production logs
Used to analyze bottlenecks, test schedule options, and plan changeovers
Helps answer questions such as:
– How can we recover capacity without new machines
– What sequence or staffing pattern best meets this mix
– How do we reduce the “delay tax” that shows up as late orders and premium freight

Most small and mid-market manufacturers can realistically reach this level within 6–12 months, if they select use cases carefully.

Stage 3: Organization-wide and Business Model Twin

At this stage, the twin is part of how the business earns revenue and shares risk.

Product as a Service, Robotics as a Service, or uptime guarantee contracts
Predictive service subscriptions based on real-time performance
Usage-based billing where the twin records utilization and performance

This level requires more organizational change, legal and commercial design, and deeper integration. For many small and mid-market manufacturers, the right approach is to understand this direction and mention it in strategy conversations, but focus execution on Stage 1 and Stage 2 use cases that deliver returns inside 6–9 months.

Why Manufacturers Must Be Use Case Driven

Most manufacturers share the same constraints: Limited bandwidth; Fragmented systems; Stretched workforce; ROI.

A use case first approach means you start where three conditions intersect:

A painful business question that matters right now
A reasonable path to data and modeling using tools and information you already have or can obtain without major infrastructure projects
A measurable outcome within 6–9 months and no later than 12

Examples that meet this bar:

Automation and cell design
Use a simulation twin to de-risk a new robot cell or packaging line. The target is fewer commissioning days, fewer change orders, and smoother client handover.
Bottleneck line operations
Create a performance twin of a bottleneck workcenter or line and test schedules, batches, and staffing patterns before touching the live system. Advanced planning tools such as Phantasma.global can complement this work.
New Product Introduction (NPI)
Build a process twin for the first 6–12 weeks of production for a new product. Model routings, cycle times, staffing, changeover rules, and WIP limits. A simple model that captures planned routing and approximate cycle times can expose unrealistic assumptions before first manufacturing cut. For a mid-market manufacturer, that can be the difference between a calm launch and a three-month scramble.

Where Manufacturers Already Pay for “As a Service”

If digital twins are eventually going to support “as a service” revenue models, it helps to start with where manufacturers are already comfortable paying recurring fees for digital capabilities. Today, it is visible in at least three areas:

Monitoring OEE and machine performance
Using computer vision for quality inspection as a service
Cloud-based ERP and MES for production, traceability, and planning

In each case, the pattern is consistent:

Clear operational pain
Limited in-house capability to build and maintain the solution
Fast, visible payback
Willingness to treat the service as operating expense, not a one-time project

Visibility as Service OEE and Machine Monitoring	Quality as a Service Computer Vision	System of Record as a Service Cloud ERP and MES
Manufacturers now pay monthly or annual subscriptions for OEE and machine monitoring tools. These systems tap into CNC controls, track uptime and downtime, and provide dashboards and daily reports. Even a small improvement in spindle time or throughput more than covers the subscription.	Manufacturers introducing AI-based visual inspection already pay for it as a subscription service, quality as a service. They upload images, label defects, and deploy models to cameras on the line. The vendor maintains the platform and the models. A subscription model allows them to start small and scale without a large upfront investment.	Cloud ERP and MES are now standard. They pay per user, per month, for order management, inventory, production tracking, and quality records. Leaders accept this because it enforces some process standardization. The value is concrete: better traceability, fewer stock-outs, fewer data entry errors, and improved planning.
Today, this is visibility as a service. Over time, the same data streams can form the backbone of an operational twin of the shop, enabling scenario planning and performance-linked contracts in specific relationships. A shout-out to my good friend Srihari at LeanWorx, who writes entertaining and relevant posts on this topic.	The image and defect data, combined with process parameters, can become part of a product and process twin for critical SKUs, supporting traceability, root cause analysis, and even warranty and insurance discussions. Quality Assurance as a Service is a natural evolution in the manufacturing ecosystem.	These systems are, in effect, system-of-record as a service. As more operational and equipment data feed into them, they become the data spine for operational digital twins of lines, plants, or even supply chains. Vendors and partners can then layer planning as a service or capacity-analysis as a service using twin-like models. (Phantasma.global)

Why XaaS and Twins Are Converging

The move toward “as a service” in manufacturing is not only about pricing. It reflects real pressure in three areas.

People
Leaders and planners do not have the time or skills to build and maintain OEE systems, vision models, or integrated ERPs. Their best people are tied up with daily production issues. They are willing to pay for services that remove complexity and provide ready-to-use insight. Digital twins will follow the same logic when they are positioned as a way to answer recurring questions, not as a technology project.

Process
Mix and demand change frequently, quality expectations continue to rise, and customers expect better traceability and responsiveness. Subscription services give manufacturers flexibility to start small, adjust scope, and change direction more easily than with a one-time capital project. Twin-driven services can fit into this by offering faster learning loops around capacity, quality, and launch decisions.

Technology
OEE tools, vision systems, and cloud ERP/MES are already collecting continuous data about machines, parts, processes, and events. That data is cleaned, structured, and surfaced as dashboards and reports that people actually use. Over time, these same data streams can feed more explicit twin models without asking manufacturers to start from scratch. The real convergence is that existing subscriptions are quietly building the data spine that digital twins need.

What Leadership Needs to Do Now

A manufacturing leader reading this should walk away with clear next moves for the next 6-12 months.

Choose two or three business questions that digital twins must help answer, grounded in current pain:
– De-risk a specific automation or NPI project
– Stabilize a known bottleneck area
– Improve visibility and trust with a key CM
Assign a business owner and a technical owner for one priority use case.
Define outcome metrics that matter to the business and are measurable with existing tools, for example:
– Capacity released: additional productive hours per week on the bottleneck
– Inventory turns or days of inventory (or WIP days) for the line in twin
– Overtime or rework hours reduced on that scope
Time-box the effort inside a 6–9 month window. This is not a side experiment.
Decide your position on XaaS:
– Where you will pay for service and performance instead of owning everything
– Where, if you are an OEM or integrator, you might eventually offer performance-based offerings yourself

The detailed work of making processes, data, and people “twin-ready” is its own topic, which I will address separately. For now, the essential step is this:

Stop treating digital twin as something only large enterprises can afford.
Start treating it as a disciplined way to de-risk decisions, protect scarce capital, and build options for how you create and capture value in the future.

If you would like to start this journey of building a digital twin, please write to us.

Structuring Engineering Data for Smart Manufacturing

Wed, 12 Nov 2025 06:29:18 +0000

How Engineering Data Drives Manufacturer Efficiency

The Hidden Cost of Unstructured Engineering Data

Engineering teams generate immense value and assets: product drawings, component models, configuration variants, and manufacturing instructions. I found that the basic SOPs for design data management frequently became diluted.

Standard naming conventions
Searchable metadata (material, performance, usage context)
Clear lineage of changes or reuse across projects
Links to production and ERP systems

If your CAD libraries are hard to search, or if each engineer creates their own naming logic, you are paying a tax across the product lifecycle. Here are common symptoms of poor data structure:

Long quoting cycles because prior designs are hard to retrieve
Rework caused by minor variations of already-solved problems
Inventory bloat due to unnecessary part variants
Slow onboarding of engineers who cannot easily navigate legacy designs
Maintenance delays from unclear service procedures or component lineage

These are not isolated engineering problems. They erode operational responsiveness, inflate working capital, and delay revenue.

Making Design Assets Discoverable and Usable

The first step in unlocking value from engineering data is to treat it as a product, not just project output. This means:

Tagging and Metadata: Standardize feature-level tags (function, fit, interface, material, tolerances) so components are searchable.
Version Control and Lineage: Track the evolution of parts and subassemblies. Know which project used what design, and what changes were made.
Searchable Repositories: Move beyond folder trees. Use visual search, attribute filters, or try AI-based similarity search to locate parts.
Bill of Information (BOI): Go beyond the BOM. Include manufacturability, quality, and service attributes in the design record.

When done well, this transforms a CAD file from static geometry into a connected, reusable, and monetizable asset. Your CAD asset now serves multiple stakeholders, both internal and external, and becomes the foundational data that can be used in multiple tools used in your organization. Here are a few ways in which engineering data enables other functions practically:

Capability	Engineering Data Role
Digital Twin	Feed geometry, tolerance and fit checks into virtual simulations
Predictive Maintenance	Link component history to service manuals and sensor alerts
Production scheduling	Enable cycle time estimation via process linked parts
Configure-Price-Quote	Identify reusable configurations, validate constraints

Making engineering data usable across the organization requires changes in standards, ownership, and mindset.

Design Library Stewardship: Assign owners for tagging, quality, and version control.
Process Integration: Link PLM or PDM systems with ERP, MES, and service platforms.
Training and Onboarding: Educate engineers to design with reuse and metadata in mind.
KPIs: Track reuse rate, part lineage clarity, and time-to-quote to demonstrate impact.

Engineers can take advantage of new tools/ feature sets inside CAD that are making this administrative task easier. We pay attention to $$ for financial accuracy. Accuracy and ease of design data usability has direct and indirect cost impact on the organization (rework costs, warranty costs, lost opportunity cost, onboarding & time to competency, etc.).

Call to Action: Start with your next project.

A full overhaul is not required to see impact. Start with one product family or platform:

Audit existing CAD files for redundancy and undocumented variations.
Define tagging standards for components and assemblies.
Link designs to quoting, production, or service use cases.
Demonstrate business impact via quoting time reduction, reuse rate, or inventory streamlining.

Once the value is visible, it becomes easier to scale.

Refer previous blog on Design Reusability: https://www.mtabusa.com/blogs/post/why-design-reuse-belongs-in-smart-manufacturing-playbook

Why Design Reuse Belongs in Smart Manufacturing Playbook

Tue, 28 Oct 2025 21:08:27 +0000

How Reuse Drives Utilization in OEM Engineering

Impact	Symptom	Cost Implications	Benefits from Design Reuse
Time	Long quoting and engineering cycles	Lost deals, delayed revenue	Faster time to quote improves win rate, especially in capital sales
Quality	Rework due to new design variants	Warranty claims, customer dissatisfaction	Reduces rework, support costs
Inventory	SKU proliferation	Inventory bloat, higher carrying costs	Lower BOM costs, simplified sourcing
Talent	Repetitive work, difficult onboarding	Time to competency	Frees up capacity for NPD

Critical Intervention: Embedding Reuse Principles into Design Practice

We started the discussion on design reuse. The design team and its leader understood the context of the problem through joint reviews and discussion. Actions taken:

Modularize all new designs
Tag them with consistent feature and naming conventions
Document configurations so variants could be easily understood

To ensure consistency and enable scale, we aligned tagging conventions with part feature, function, and interface fit. This made it easier to navigate the design library when seeking a component with similar constraints.

My observation: Engineers value creativity, rigor in design and dislike documentation and this change initially met with some resistance. Leadership positioned designs as long-lived assets, and their reuse directly showcased how impactful the engineers' work is in driving both throughput, quality and customer experience. We defined ‘reuse’ as sub-assembly reused with less than 10% design change. This let design leaders track module lineage across projects and identify which designs were repeatedly customized. Within six months, over 20% of common design tasks were being fulfilled using existing asset configurations. Key metrics tracked:

% increase in part library reuse
% of SKU reduction for inventory management

Digital Enablers for Reuse

Design reuse requires more than policy. It requires digital tools and CAD & ERP data hygiene. What it looked like for us:

Engineering Asset Libraries with searchable metadata, images, and part lineage
Simple product digital twins, models that simulate product configurations, helped us pre-check fitment, compatibility and interface of reused modules.

These digital enablers also provided early visibility into feasibility, and surfaced interference or manufacturability challenges that had already been resolved elsewhere in the system.

Now, tagging Assistants that auto-categorize designs based on shape, performance, or project history are available, but we did not get to use them.

What does Governance look like?

It requires and collaboration across departments:

Role Clarity: Who curates the design library? Who approves new modules?
Documentation Standards: What constitutes a “reuse-ready” design?
Reuse Metrics: How often is a design reused? By which teams? In what context?
Design Reviews: Periodic audits to identify reuse candidates and archive legacy variants

Incentives also matter. Tracking reuse can help showcase engineering effectiveness and free up cycles for new product innovation. In our case, reuse champions emerged organically.

Translating Industry 4.0 Maturity into Financial Outcomes

Tue, 12 Aug 2025 23:46:41 +0000

From shop floor metrics to strategic investment decisions

In manufacturing, the conversation between the operations team and the boardroom often breaks down when it comes to digital transformation. Operations teams talk about Industry 4.0, connected systems, and predictive analytics. The boardroom is concerned with the return on invested capital, risk-adjusted returns, and capital allocation priorities. Without a clear translation layer, investments in smart manufacturing either stall or fail to deliver the expected outcomes.

The Manufacturing Technology Balance Sheet (MTBS), grounded in structured assessments such as the Smart Industry Readiness Index (SIRI), provides that translation. It enables manufacturing leaders to connect their technology readiness (shopfloor, enterprise, organization) with financial decision-making in a way that is quantifiable, comparable, and actionable.

Why the Manufacturing Technology Balance Sheet Matters

Creates a clear baseline of digital maturity across people, processes, and technology.
Identifies operational blind spots that are not visible through traditional financial statements.
Links technology gaps to financial risk such as extended lead times, higher downtime, or inability to scale for large orders.
Enables capital allocation sequencing to avoid investing in areas where foundational readiness is lacking.

When presented in the boardroom, the MTBS acts as a companion to the financial balance sheet. One shows the organization’s fiscal health; the other shows its capacity to sustain and grow that health through technology-enabled performance.

A Sample Illustration

A mid-market electronics manufacturer conducts a MTBS review and discovers that its low maturity in equipment connectivity was inflating maintenance costs and extending lead times. By investing in a targeted machine connectivity and analytics project first, rather than jumping straight into a full MES implementation, the company reduced downtime by 18 percent, improved delivery performance, and built the operational foundation needed for future MES deployment and its successful adoption.

Practical Actions for Manufacturing Leaders

Treat mfg technology readiness as a measurable asset, not an abstract concept.
Lean on cross-functional teams that include finance, operations, and IT in maturity assessments in recommendations.
Use MTBS outputs to challenge assumptions about which projects deliver the fastest and most sustainable ROI.
Ensure every boardroom discussion on capital investment includes a technology readiness context.

Conclusion

The Manufacturing Technology Balance Sheet is more than an operations tool. It is a bridge between the language of machines and the language of money. By translating digital readiness into financial terms, manufacturing leaders can secure boardroom alignment, prioritize high-value projects, and ensure that every dollar invested accelerates both operational performance and long-term profitability.

Call to Action: Schedule an MTBS-to-financial linkage review for your organization, where operational maturity gaps are mapped directly to financial outcomes and capital allocation priorities. Position this as a decision-support step before the next budget cycle.

Why Manufacturing Needs Digital Fluency Before AI-Native Talent

Thu, 17 Jul 2025 05:06:57 +0000

Beyond the Buzzword

Three Imperatives for Manufacturers and Career Influencers:

1. Begin Before the Career Begins: Awareness at the School Level

The manufacturing AI-native workforce cannot be built at the point of hire. Manufacturers must invest in community awareness campaigns, STEM-to-shop-floor pathways, and hands-on engagement starting in middle school. Early exposure drives relevance and aspiration.

Partner with local schools for curriculum modules on modern manufacturing
Offer facility tours, job shadowing, and virtual experiences
Use digital tools (VR, gamification, AR) to demystify factory jobs
One of my favorite shows was "How It's Made"

2. Shift from AI Skills to Digital Fluency and Data Literacy

AI in manufacturing is not about writing code. It is about applying and interpreting dashboards, responding to alerts, troubleshooting machines, and making decisions based on contextualized data.

McKinsey notes that manufacturers face up to 800% productivity variation based on skill proficiency. Closing this gap requires:

Training in digital & tablet-based workflows and MES interfaces on the factory floor
Partnering learners (trades people and engineers) with AI for better outcomes
Teaching frontline teams to recognize, tag, and escalate digital signals
Embedding data interpretation into daily roles, not just analytics teams

3. Institutionalize Knowledge Capture Through Technology and Culture

The brain drain from retirement is real. Manufacturers must make knowledge capture a built-in behavior, not an IT initiative.

Use connected worker platforms to embed SOPs, videos, and peer insights
Assign responsibility for documenting tribal knowledge at every role level
Integrate video, sensors, and wearable data into workflows

One aerospace firm cited by McKinsey increased throughput 15% by pairing less skilled staff with a veteran and embedding documentation in their shift routines- losing productivity in the short term but building capacity and capability of the workforce in the long term.

Conclusion:

The path to a truly “AI-native” manufacturing environment lies not in waiting for digital evolution to occur naturally, but in actively crafting a digitally fluent workforce. Early educational outreach, strategic upskilling, and robust knowledge management systems are the cornerstones of this transition. Manufacturers who adopt these measures will not only overcome the talent gap but will also drive efficiency, innovation, and sustainable growth.

In manufacturing ecosystem, the infrastructure for “AI-native workforce” does not yet exist in most factories. Manufacturers must first cultivate digital-first mindsets—where every worker is aware, fluent, and connected. We lay the groundwork for them to grow into Industrial AI future, making today’s manufacturing resilient, attractive, and scalable.

References: McKinsey Insights on Workforce Digital Transformation published Jul 2, 2025

Get Started Now

Why Building with AI in Manufacturing Must Keep Humans in the Loop

Mon, 16 Jun 2025 05:07:49 +0000

Building with AI in Manufacturing

Here’s why a human-centered approach to building with AI is not optional—it is essential.

🔹 AI Is Becoming as Foundational as the Internet—But Requires Human Context

Just as the internet is now inseparable from how we work and live, AI is moving toward becoming a virtual assistant for performance—augmenting how we make decisions, optimize workflows, and design products. In engineering tech, firms like Siemens, PTC, Dassault, and Autodesk and startups (BaseTwo AI, Quarter20, MetAI) are leading this evolution with digital-native and AI-native product development workflows. The digital twin, once futuristic, is now a practical tool for virtually testing and validating designs before a single tool cut is made.

This is transforming not only how capital is deployed but also business models themselves—from factory-first to digital-first, and from asset-heavy to production-as-a-service. This is giving rise to digital & AI first hardware companies and their capital requirements and supply chain approach are different. But no matter how advanced the tech, humans are still essential to closing the loop between simulation and execution.

🔹 The Shop Floor Is Not a Sandbox—Fear, Legacy Systems, and Fragmented Data Persist

Despite AI’s promise, end users in manufacturing face real barriers:

Fear of job loss remains high. Any mention of automation or AI immediately raises questions about livelihood, particularly among frontline workers.
Many factories still run on 1980s and '90s equipment, making integration with modern AI tools complex and expensive.
Cybersecurity gaps are widening as legacy infrastructure meets connected systems—ransomware attacks are growing year over year.
Data is fragmented and unstructured, spread across CAD files, paper notes, ERP, CRM, and maintenance logs.
There is a critical dual skills gap: aging workforce & its tribal knowledge and attracting AI-savvy talent with domain fluency?

AI cannot succeed in manufacturing unless it addresses these human and operational realities.

🔹 Change Management Must Be Deeply People-Centric

Change is hard—especially in manufacturing, where culture and legacy practices are deeply embedded. Success in implementing smart factory technologies requires engaging everyone, not just leadership.

From my own experience, adoption accelerates when:

Grassroots workers are included early in discussions and training on new technologies.
Engagement spans the entire factory value chain: plant heads, supervisors, operators, maintenance, IT, and HR teams.
Each role understands not only the tools but also the value AI brings to their specific job function.

In the B2B industrial manufacturing sector, the challenges of AI adoption are often shared between the supplier and the customer. Interestingly, the solutions you develop internally—for your workforce and factory—can often be turned into assets that enhance customer experience and deepen supplier partnerships.

Globally, manufacturing is a heavily regulated industry, and regulatory bodies are still adapting to the pace of AI transformation. In one instance, I encountered a factory where digital systems offered greater traceability and data integrity for workforce protection, yet the labor authority continued to require paper logbooks with handwritten signatures. This underscores the urgent need for policy modernization to align with technological advancements. On the product front, I have worked with a startup, Saphira AI, which applies AI in safety compliance. Their approach has the potential to help product manufacturers accelerate certification readiness by integrating AI-driven safety and quality checkpoints into their production workflows. Another example: Computer vision, for example, is increasingly being accepted in real-time safety monitoring and traceability applications. These tools illustrate how human-AI collaboration can enhance both productivity and worker protection.

Still, the tension between privacy and utility persists. As AI continues to expand its role on the factory floor, ethical questions around surveillance, consent, and data ownership will remain at the forefront—demanding thoughtful, people-first leadership.

🔁 AI Adoption Isn't Just a Technical Rollout—It is an Organizational Shift

To succeed, AI must:

Support and extend human capability, not attempt to replace it.
Be embedded in existing workflows, not layered on top of them.
Include feedback loops so workers can improve, adjust, and trust the system.
Prioritize skills development, role evolution, and cultural alignment.

AI in manufacturing will scale not through force, but through collaborative intelligence—where every person, from engineer to operator, becomes a part of the transformation story. Manufacturers, especially small and mid-scale, should assess their organization's digital maturity, critical processes to be automated and skill gaps and build their plan to leverage the emerging tools. For small and mid-sized manufacturers, this means starting with a clear-eyed assessment of their digital maturity, identifying the critical processes ripe for automation, and addressing the skill gaps across their workforce. With this foundation, they can strategically adopt emerging AI tools to respond to opportunities and uncertainties with .

Get Started Now

Blueprint to Building Blocks: Accelerating Engineering Design Through Reuse via Workflows and AI

Tue, 03 Jun 2025 05:20:37 +0000

Practical Tips on How to Reframe Design Process and Thinking in Manufacturing

My goal in this article is to identify a few practical ways design engineers can incorporate reuse and automation in their design process. I come at it from a place of curiosity and how I worked with my team. Design is a critical pillar of digital transformation—not only as a contributor, but as a function that must itself evolve. Transforming the design process through automation and AI is essential to enabling agility and integration across the entire manufacturing value chain

Manufacturers, OEMs and others, are under pressure to deliver customized products faster, at lower cost, and with greater consistency. Yet in many engineering departments, tribal knowledge, limited design reuse, and underutilized tools are slowing down innovation. We frequently see repeated effort, SKU proliferation, and inconsistent product quality.

Design reuse is not just about saving time. It is a strategic lever for improving engineering productivity, ensuring design consistency, and enabling scalable digital transformation. Think of this as a circular economy principle applied to product design- extend design asset life, responsible (re)sourcing, repurposing!

The Hidden Cost of “Starting from Scratch"

Many design departments struggle with:

Poor searchability of past design files, drawings, or BOMs.
Lack of modular design templates for commonly used sub-assemblies
Missing integration between mechanical, electrical, and control engineering
Inconsistent annotations or design history that limit reuse and collaboration

I recall how challenging onboarding a new engineer was in our teams due to the above issues and how they showed up in the manufacturing and assembly process, resulting in rework, lost time and cost.

Cross-Domain Collaboration

Integrate MCAD and ECAD platforms for electro-mechanical builds.
Use simulation to validate design fitment and assembly logic before build. Fitment was one of the biggest challenges and issues cropped up during prototype fabrication. Fitment features/ analysis are underutilized.

Searchable, Annotated Digital Twins:

Your drawings are your entry-level digital twins
Transform static 2D drawings into annotated 3D models with tribal insights, design constraints, and change history embedded
Use PDM/PLM systems (e.g., Siemens Teamcenter, PTC Windchill, Dassault ENOVIA) to version, track, and audit reuse metrics

🔍 Industrial Foundation Models: The New Assistant on the Design Desk

The rise of Industrial Foundation Models (IFMs)—large-scale AI models trained on multi-modal engineering and manufacturing data—is changing how design teams work. These models offer contextual intelligence, helping engineers reuse designs more confidently, test alternatives faster, and reduce errors earlier.

Real-World Application: SolidWorks Workflow for Annotating a Robot Arm Base

Here is one possible workflow to turn an existing mechanical assembly (e.g., a robot arm base with motors and cycloidal gears) into a smart, annotated asset in SolidWorks:

Prepare the Assembly: Organize and constrain your .SLDASM file with configurations.
Use DimXpert and Model Items: Auto-apply manufacturing dimensions and tolerances.
Feature Recognition with FeatureWorks: Recover editable features from legacy parts.
Enhance with 3DEXPERIENCE AI: Use AI to suggest part reuse and missing metadata.
Publish with MBD: Export annotated 3D PDFs for collaboration.
Archive in SolidWorks PDM: Tag, search, and reuse designs across teams.

Sources: SolidWorks Help, FeatureWorks Guide, MBD White Paper, Dassault 3DEXPERIENCE AI Docs.

Tangible results:

Albeit not fully automated, at MTAB Engineers, we changed our design process. We incorporated modular thinking of design and clear annotation in our design work. It helped us significantly reuse design assets and create platforms for robotics, mobile robots and CNC machines:

Speeding up our time to market by 25-30%;
Understanding impact of substitute components and extent of changes;
Creating virtual simulations for customer specific applications;
Most importantly, keeping internal and external documentation up to date.

Closing:

In a world of increasing complexity and convergence of IT and OT, design reuse isn’t a shortcut—it is a competitive advantage. When OEMs and parts manufacturers structure their product design around reusability, engineering becomes a productivity engine rather than a bottleneck.

If you would like a white paper on this topic, drop me a note.

Part 3: Human-Centered Intelligence: How Digital Twins Empower the Connected Worker in OEM Manufacturing

Tue, 20 May 2025 05:10:57 +0000

Series: Designing Intelligence – The Strategic Use of Digital Twins in OEM Manufacturing

Introduction to the Series

In this three-part series, we explore how digital twins can help OEM manufacturers move from reactive decision-making to intelligent, data-driven operations. From early-stage design to procurement strategy and workforce enablement, digital twins—when implemented beyond surface-level 3D models—can serve as an integrated, cross-functional foundation for growth and resilience. The discussion is focused towards OEM in industrial engineered goods segments.

This three-part series explores how digital twins can serve as a force multiplier across:

Part 1: Design agility and engineering integration
Part 2: Operational Digital Twins for shopfloor and supply chain management
Part 3: Connected worker enablement and training scalability

1. Capturing and Transferring Tribal Knowledge

In OEM manufacturing, different roles—from assembly technicians and field service engineers to calibration teams and quality inspectors—rely on decades of hard-earned knowledge. Much of what makes production successful is invisible: tribal knowledge passed from veteran operators through informal conversations, trial-and-error, and years of repetition. At our factory, one of our 25+ years' tenured production supervisor could take one look at mechanical issue, know exactly what caused it and how to resolve it. The problem? This knowledge does not scale, and it often walks out the door.

Digital Twins can be used as assets to showcase how shopfloor workforce and production executives build repeatably through Process, Practice, Consistency and Problem-Solving. When thoughtfully designed, digital twins help close that gap by turning unspoken knowledge into a structured, accessible format that supports every role on the floor. These tools allow teams to:

Embed operational insights directly into process simulations. Today's AI-based vision systems can watch 60–100 assembly builds and automatically identify the sequence of work instructions, tools needed, and the problem areas to watch out. (Most recently, I saw an interesting demo of Retrocausal ).
The older-hands can annotate procedures with warnings, tips, or alternative approaches (Quarter20 has been doing some fantastic work by integrating with CAD).
Create shared, visual references accessible to all shifts and facilities
Enable workers to contribute feedback or updates to the twin—flagging steps that cause confusion or proposing improvements. This keeps the digital twin relevant and collaborative.

This isn’t just about documentation. It is about transforming unstructured human know-how into a repeatable, shareable, and upgradable operational asset.

2. Scenario-Based Training and Virtual Walkthroughs

For most next-gen talent and the career influencers in their lives, manufacturing still evokes the picture of labor-intensive, grease-filled work. To talent entering the workforce, the manufacturing floor can be daunting, rigid, and overly structured. Using digital twins and immersive tools, we can address these concerns and educate them on the meaning behind these procedures and standards—how they ensure safety, quality, and reliability in the physical products we all rely on.

Pairing digital twins with AR/VR, tablet-based interfaces or dashboards enables immersive, scenario-based training. Workers can interact with real-world simulations of compliance & safety, complex assemblies, changeovers, and service routines before ever touching the line. (Check out DeepSight andtheir work in compliance and scenario-based training.)

The real value? These environments allow workers to make mistakes without disastrous consequence, reflect on what went wrong, and reinforce correct procedures in a safe way. This builds confidence and instills a deeper understanding of why each step matters—critical for safe operation and consistent quality. OEM equipment is often highly integrated, combining mechanical, electrical, and programmable elements. Training through digital twins helps workers develop practical fluency with these systems.

Because these simulations reflect actual equipment, layouts, and product variants, training becomes contextual, not theoretical. Workers are not just memorizing steps—they are developing judgment and muscle memory to perform effectively on the floor.

3. Real-Time Decision Support on the Floor

A digital twin isn’t just for engineers—it is a decision support platform tailored to each role on the floor.

Assembly technicians can check build instructions that update with the product variant.
Quality inspectors can visualize tolerance limits and recent deviations.
Field service teams can simulate a fault scenario before going on-site.
Maintenance techs can anticipate required interventions based on digital logs.

OEM environments often run on a mix of legacy equipment, PLC-based systems, and modern controls. Digital twins serve as a layer that brings these elements together, helping workers make sense of system-level interactions and operate more effectively.

When connected workers can access a real-time view of operations through one of the many gadgets, they gain:

Live machine status and material availability
Context-sensitive work instructions and tools
Alerts tied to current part mix, shift, or customer requirements

This kind of support helps frontline teams move from simply executing tasks to actively solving problems. It reduces escalations, increases throughput, and empowers workers to operate with confidence.

It also allows for faster alignment with engineering when last-minute changes occur. Instead of waiting for clarification, workers can see the updates directly in the twin—with full traceability and impact context.

4. Building Engagement and Retaining Talent

Digital enablement is not just about output—it is about people.

Frontline workers today, especially Gen Z and younger millennials, expect digital continuity between their social lives and work environments. They want tools that are intuitive, mobile, and responsive. They expect transparency, feedback loops, and autonomy.

They also want to spend less time on administrative tasks. Automated data capture, workflow automation, and GenAI-powered assistants can significantly reduce paperwork and help them focus on problem-solving and collaboration.

Digital twins can deliver on those expectations by:

Offering real-time guidance and troubleshooting tools
Enabling remote access to experts for faster problem-solving
Automating repetitive tasks, data capture, and form completion
Creating role-specific dashboards and mobile interfaces that adapt to each user’s needs

However, digital tools alone do not guarantee adoption. For many OEMs, the shift to connected systems requires thoughtful change management. Veteran workers may be skeptical, while new employees might need coaching. Success depends on cross-generational mentoring, training, and workforce members who model adoption.

Many OEMs also operate with fragmented systems—legacy ERP, standalone training platforms, and homegrown MES solutions. The good news? Digital twins can layer onto these systems using APIs and data connectors. This approach allows organizations to build connectivity step by step, without needing to overhaul everything at once. (Note: Sometimes, you do need to rip out that existing system and put in a new one, due to too much baggage.)

This kind of empowerment directly impacts retention. According to McKinsey, it costs $52,000 to replace a single frontline manufacturing worker who leaves. Reducing churn by even a small margin translates to major gains in cost and continuity. Better training through digital twins can also shorten onboarding by up to 30–40%, and reduce early-stage production errors—directly impacting efficiency and quality.

OEMs can measure success using metrics like:

Time to proficiency for new hires
Reduction in rework or quality escapes
Number of escalations resolved at the operator level
Utilization of training modules or knowledge contributions

Source: From hire to inspire: Getting and keeping Gen Z in manufacturing

5. New Revenue Opportunities from the Connected Worker Strategy

What OEMs implement for their internal processes and people can become valuable differentiators in the products and services they offer customers. When digital twins are used to eliminate non-value-adding tasks, automate repetitive processes, and improve real-time decision-making, these capabilities can be embedded as features into the equipment or bundled as value-added services.

The challenges OEMs face internally—such as improving visibility, increasing flexibility, and addressing workforce skill gaps—are often mirrored by their customers. The systems and tools OEMs build to support their own connected workforce can be productized and extended to clients who are also seeking to modernize their factory operations.

For example:

A digital twin originally used for operator guidance and maintenance tracking can become a customer-facing application.
Knowledge capture platforms for tribal wisdom can be integrated into after-sales support systems.
Real-time dashboards used by OEM line leads can be repackaged as part of remote monitoring solutions for clients.

This creates a dual benefit:

- OEMs improve internal performance and workforce capability.

- These investments open new commercial pathways through service contracts, digital subscriptions, or premium equipment options.

Workers, who feel supported and valued are more likely to stay, grow, and lead—strengthening the factory’s long-term capability. This win-win scenario creates a flywheel effect of developing human capital and smart tools across the manufacturing value chain.

Conclusion: People, Not Just Processes

Digital twins started as tools to model machines, but their greatest impact may be in modeling and supporting people. In the context of the Connected Worker strategy, digital twins become co-pilots for the modern factory operator: preserving knowledge, guiding action, and fostering a culture of learning.

If the first wave of Industry 4.0 was about systems, the next wave is about humans. And the digital twin brings them together.

Building a Connected Worker strategy is a necessity driven by the changing demographic profile of the workforce, adoption of manufacturing technology solutions and the flexibility required to meet customer requirements and preferences.

Thank you for following the series. If you missed Part 1 (Design-Side Intelligence) or Part 2 (Operational Digital Twin), they are linked here.