In my previous blog outlining the adoption of cost engineering, I explored the dynamics behind the market move away from sole reliance on traditional, backward-looking cost estimating to one that also incorporates modern “should-cost” methods. The reasons are many, of course, but it is clear that industrial organizations are keen to use AI-driven methods and other digital tools to build much stronger layers of resilience and competitive advantage necessary to compete in today’s hyperconnected economies.

Although digitally enabled results can sometimes be achieved in an operational vacuum, digital maturity cannot. The former can demonstrate benefits like efficiency, cost reduction, safety, etc., but it will rarely scale. The latter delivers market success via competitive excellence, providing a means for better organizing the business and orchestrating the ecosystem to anticipate and meet modern market signals.

Modernizing the supply chain is, at its core, a human-centered endeavor. The successful integration of cost engineering demands significant realignment and reskilling of people. As I began discussing almost a decade ago, the workforce transformation required to modernize is certainly the most difficult endeavor a business will face.

In this blog, I’ll dive into the human element of cost engineering. I’ll touch on how roles and attendant knowledge, skills, and abilities (KSAs) across the supply chain are evolving, discuss the cultural hurdles organizations must navigate, and outline how companies can transform traditional estimators into strategic consultants.

Tribal Knowledge: I Feel Like I’ve Been Here Before

Leadership must address the workforce crisis currently confronting industrial manufacturing. Look at any credible information resource and the numbers are basically the same. Whole industries are facing rapid workforce retirements, with approximately 25 percent of the total manufacturing workforce already over the age of 55. Within small and medium-sized enterprises, which form the bedrock of the industrial manufacturing supply base, particularly in North America, between 30 and 40 percent of business owners and skilled operational workers are nearing retirement age. Ouch.

And yet we’ve known this has been underway for quite some time, but here we are. Historically, the reaction to tribal knowledge was wariness. I recall many conversations with leadership and frontline workers as technologies such as machine learning were initially deployed. Tribal knowledge, expertise, and the workforce that owned it were often treated as a nut to be cracked and the insides taken. Initially, the shell was perceived to be obstinately hard, with workers guarding their critical expertise, including core intellectual property (IP), as a means of fending off obsolescence. It didn’t lend itself to, shall we say, everyone pulling in the same direction.

Supply chain was no exception to this pattern. Cost estimating relied heavily on the undocumented tribal knowledge and personal experience of veteran employees. As these experts exit the workforce, they take decades of specialized intuition with them, leaving organizations highly vulnerable.

As a result, a new discipline has taken hold, as tribal knowledge is likely to be unretrievable in many instances or, in situations where leaders show a lack of humility, downsized too quickly. Modern cost engineering takes aim squarely at the reliance on human memory with standardized, process-based cost models and empirical data. Yet, an overwhelming 90 percent of supply chain leaders report a severe lack of the digital talent required to operate these new systems. Here we are, again, back to the ever-important human element at the center of a technology endeavor.

Redefining Supply Chain Personas

Rather than taking the same, lose-lose historical approach to cracking tribal knowledge, leading organizations are pivoting workers away from the manual, unsafe, and repetitive. What they are doing differently, though, is concertedly moving subject matter experts toward higher-level orchestration and critical oversight. It won’t pan out with every worker, certainly, but it will ensure that the expertise is retained and applied to creating more strategic value. On the surface, that presents much more opportunity for a win-win scenario. Here is how some specific roles are evolving:

Estimator

Historically, manufacturing estimators spent most of their time immersed in manual, backward-looking work. They pored over static 2D PDFs, visually interpreted complex 3D CAD models, and stitched together cost assumptions from disconnected spreadsheets. Much of their value came from patience and pattern recognition rather than insight, and the process was slow, reactive, and highly dependent on individual experience. For leading companies that are aggressively implementing cost engineering processes, that is radically changing.

In the world of cost engineering, this role is now that of a strategic advisor. Leveraging AI to automate much of the data extraction that once consumed their time, this role develops models to identify cost drivers based on real manufacturing constraints and material behavior. As a result, this role now focuses more on guiding internal teams on design-for-manufacturability decisions and outlining strategic trade-offs that can include a mix of potential metrics, such as cost, lead time, and, increasingly, carbon impact.

Procurement

Procurement has primarily been about transactional efficiency and negotiation. Success was generally determined by price, often with significant visibility limitations into how the price was constructed. Framed within cost engineering, procurement is driven by collaboration and risk management. Using precise cost models, sourcing conversations begin with a clear understanding of cost, informed by specifics on materials, labor, processes, and capacity constraints. If a supplier’s quote exceeds cost expectations, conversations can then be had specifically about how to target specific constraints, such as inefficiencies in process or materials. The objective is to provide transparency that allows for a win-win relationship in terms of performance, profitability, and reliability.

Frontline

Despite the best of intentions to change the reactive nature of the role, frontline work has been dominated by manual execution and post-problem decision-making. Operators were tasked with keeping machines running, responding to breakdowns as they occurred, and relying heavily on tribal knowledge passed down informally and gained over time. Cost engineering shifts the dynamic for frontline workers. Upstream processes and systems provide precision that is communicated to these workers in terms of production expectations. Operators are tasked with supervising processes, identifying deviations, and capturing machine-level issues as they occur. As these workers become more connected and augmented via technology, faults and anomalies are logged digitally, with automated routing to maintenance or engineering as needed. With effective cost engineering, the frontline workforce ensures production aligns with cost and performance expectations.

Chief Supply Chain Officer (CSCO)

In the past, supply chain leadership was back-office oriented, using historical information to attempt to optimize logistics execution, inventory control, and cost. Their influence was significant but fairly tactical. That orientation shifts significantly with cost engineering as the CSCO becomes the central orchestrator of enterprise performance, based on the organization’s ability to align with market demand. Supply chain data increasingly impacts revenue and margin stability, based on market responsiveness. As a result, the CSCO sits at the intersection of strategy, technology, and execution, with an increased mandate that expands beyond moving goods to shaping how the organization makes decisions. In an organization using cost engineering, CSCOs are redesigning roles, workflows, and governance models, based on AI-driven insights that orchestrate decision-making across the enterprise and ecosystem.

Aversion to Change: You Can’t Take the Human Out of, Well, the Human

So, implementing cost engineering seems like an obvious win. Despite the obvious operational benefits, integrating cost engineering introduces complex modernization challenges. Of course, these challenges are mostly rooted in aversion to change. It’s a pretty understandable problem, with generations of workers having been trained on historically based methods and having spent entire careers honing a requisite expertise. To them, AI and automated decision-making are met with deep suspicion, rightfully grounded in the fear that technology will replace jobs and render their expertise irrelevant. They are not wrong. This challenge has been exacerbated by leadership deploying complex new software without context. In reaction to these poorly orchestrated, technology-centric changes, operators bypass the systems and revert to familiar methods and tools, neutralizing investment and anticipated benefits. Pilot purgatory, anyone?

To counter this within the organization, leadership must employ empathy, transparency of intent, continuous learning, and AI explainability that enables humans to trust machines and the logic behind their decisions. From an external perspective, organizations also need to understand that they are only as strong as their weakest supplier. Leading companies gain their status by subsidizing the digital and cybersecurity capabilities of their ecosystem. It becomes a case of a rising tide lifting all boats.

Return of Value

Deploying cost engineering cannot be about eliminating the human workforce through automation. It relies on a human-on-the-loop model, but it defers to technology to manage massive data complexity. The role of expert workers is to apply contextual judgment and engage in continual collaboration. The transition to this approach requires transparency and significant digital upskilling that will likely feel uncomfortable initially. Due to the step change required in this shift, organizations need to define and align with a return of value rather than shorter-term return on investment. By empowering the workforce and supply chain ecosystem to employ data-driven precision, the organization transitions from a guesswork culture to one of definable competitive differentiation.

In blog three of this series, I’ll explore the process component of the equation. I’ll focus on departmental silos, cross-functional teams, and supply chain orchestration. You can read the first blog in this four-part series here.

The post Modern Cost Engineering Evolution: Rewiring the Human Element for Supply Chain Resilience appeared first on Logistics Viewpoints.

The post Amazon Launches “Supply Chain Services” Leveraging its Global Logistics Network appeared first on Logistics Viewpoints.

The next phase of supply chain AI will not be defined by better models alone. It will be defined by whether those models can improve real decisions across planning, logistics, sourcing, fulfillment, and risk management.

Download the full ARC Advisory Group white paper, AI in the Supply Chain: From Architecture to Execution, for a deeper framework on how supply chain AI is moving from technical architecture toward decision intelligence, operational execution, and coordinated action across planning, logistics, sourcing, fulfillment, and risk management..

Artificial intelligence has moved quickly through the supply chain conversation.

The first wave focused on what AI could do. Could it improve forecasts? Detect disruptions? Summarize documents? Support planners, buyers, dispatchers, and customer service teams?

Those were useful questions. They helped establish the architectural foundation for AI-enabled supply chains: agent-to-agent communication, retrieval-augmented generation, graph-based reasoning, persistent context, and more interoperable data environments.

But architecture is not execution.

The harder question now is whether AI can improve real operating decisions inside complex supply chains. These are decisions involving cost, service, inventory, capacity, risk, customer commitments, physical assets, and financial consequences.

A model may forecast demand. A visibility platform may detect a disruption. An agent may recommend a response. None of that matters much unless the organization can turn the signal into coordinated action.

That is the focus of this new Logistics Viewpoints series.

From AI Capability to Operational Decision-Making

The first phase of supply chain AI was about capability. The next phase is about consequence.

Supply chains are not abstract information systems. They are physical operating networks. A transportation decision changes cost and service. An inventory decision affects availability and working capital. A sourcing decision changes risk exposure. A warehouse decision changes labor, throughput, and customer performance.

This is where many AI programs stall.

They produce insight, but the workflow does not change. They generate recommendations, but decision ownership remains unclear. They detect exceptions, but the organization still responds through manual handoffs, email chains, spreadsheets, and delayed escalation.

The result is decision latency: the gap between when a condition changes and when the organization executes a coordinated response.

In volatile supply chain environments, decision latency is not just an inconvenience. It becomes a structural weakness.

Why the Decision Intelligence Layer Matters

Enterprise supply chain technology has long been organized around systems of record and systems of planning.

ERP, WMS, TMS, order management, procurement, and planning platforms remain essential. They preserve transactions, manage workflows, and support structured planning processes.

AI introduces the need for another layer: a decision intelligence layer.

This layer does not replace existing systems. It operates across them. It connects signals, context, reasoning, governance, and execution. It helps the enterprise evaluate conditions continuously, understand tradeoffs, and support or initiate action within defined boundaries.

That distinction matters.

Not every AI system should be allowed to operate near physical or financial consequence. The closer AI gets to execution, the greater the need for context, determinism, governance, auditability, and human oversight.

Supply chain AI is not one category. It is a set of capabilities that must be matched to the decision environment in which they operate.

What the Series Will Cover

This ten-part series examines how supply chain AI moves from technical architecture to operational execution.

The series will cover:

1. From Capability to Execution
Why the supply chain AI conversation is moving beyond pilots, demonstrations, and technical capability toward measurable operational impact.

2. The Decision Bottleneck
How fragmented systems, functional handoffs, and delayed escalation create decision latency across modern supply chains.

3. From Systems of Record to Systems of Decision
Why AI adds a new decision layer above ERP, planning, TMS, WMS, and visibility platforms.

4. Operational AI Requires Action Pathways
Why AI insight has limited value unless it connects to workflows, owners, thresholds, execution systems, and feedback loops.

5. Five Requirements for Operational AI
The operating requirements that separate useful AI from AI theater: decision-ready data, contextual intelligence, action pathways, governance, and closed-loop learning.

6. From Agent Communication to Coordinated Execution
Why agentic AI matters only if it improves cross-functional coordination, not simply because agents can communicate.

7. Context Becomes a Requirement
Why supply chain AI must understand history, supplier performance, customer commitments, contracts, network dependencies, and prior exceptions.

8. Planning and Execution Are Converging
How AI changes the cadence of supply chain management by embedding planning logic inside execution workflows.

9. Market Structure: From Functional Software to Decision Architectures
Why buyers should increasingly evaluate technology providers by the decisions they improve, not only by the software category they occupy.

10. Operating Model Implications
How decision-centric AI changes roles, metrics, governance, accountability, and the future work of supply chain planners and operators.

The Buyer Question Is Changing

For years, supply chain technology evaluation has often started with functional categories.

What does the system do? Is it a planning platform, TMS, WMS, visibility solution, risk platform, procurement tool, or analytics application?

That question still matters. But it is no longer sufficient.

The more important question is becoming: what decisions does this system improve?

Does it improve replenishment decisions? Transportation decisions? Supplier risk decisions? Inventory allocation decisions? Customer commitment decisions? Exception resolution decisions?

And just as important: how does the recommendation connect to execution?

This is where the market is moving. Planning vendors, execution platforms, visibility providers, risk intelligence solutions, and enterprise software companies are all embedding AI more deeply into their offerings. Their starting points differ, but the direction is consistent.

The market is shifting from functional software toward decision-centric architectures.

That shift will create opportunity, confusion, and new evaluation challenges for buyers.

Why This Matters Now

Supply chain leaders are not short on AI claims.

They are short on proof.

They need to know where AI can improve real decisions, where it should remain advisory, where autonomy is inappropriate, and where governance needs to be built before scale.

They also need a practical way to separate serious operational AI from generic AI positioning.

That requires a more disciplined conversation. Not just about models. Not just about agents. Not just about data. But about decision environments, operating consequences, and the architecture required to move from insight to action.

Closing CTA

Logistics Viewpoints and ARC Advisory Group are examining how decision intelligence, agentic AI, contextual reasoning, and next-generation supply chain architectures are reshaping supply chain technology markets.

Follow this ten-part series on Logistics Viewpoints as we examine how supply chain AI is moving from architecture to execution.

To discuss how these changes may affect your organization, schedule a 15-minute analyst conversation with Logistics Viewpoints / ARC.

We will listen to your situation, offer a candid outside perspective, and, where appropriate, suggest practical next steps or areas where ARC research and advisory support may help.

The post AI in the Supply Chain: From Architecture to Execution appeared first on Logistics Viewpoints.