Speak at ARC 2026: Prove AI Supply Chain ROI

December is when most supply chain teams do two things at once: close the books on peak season and quietly rewrite next year’s risk plan. If you’re in transportation or logistics, you already know why. Disruption didn’t “go away” in 2025—it just changed shape. Geopolitics shifted lanes, labor stayed tight, carriers kept repricing volatility, and customer expectations didn’t soften.

Here’s the stance I’ll take: the industry doesn’t need more AI demos. It needs more proof. Proof that AI improves on-time delivery, inventory turns, supplier performance, and cost-to-serve—without creating a fragile, black-box operation.

That’s why the ARC Industry Forum 2026 call for speakers matters for anyone working on AI in supply chain & procurement. This is one of the few stages where practitioners can share what worked, what broke, and what they’d do differently—so other teams can make better decisions faster.

If you’ve deployed AI in transportation or logistics, your story is more useful than your slide deck.

Why ARC 2026 matters for AI in supply chain & procurement

Answer first: ARC is valuable because it rewards operational truth—real deployments, measurable outcomes, and practical frameworks—over vendor theater.

The forum’s focus on intelligent operations and resilient supply chains lines up exactly with the problems AI is being asked to solve right now:

• Demand volatility and forecast error that breaks S&OP
• Network disruptions that require rapid re-optimization (not quarterly redesign)
• Supplier risk that can’t be managed with questionnaires alone
• Warehouse and transportation execution gaps caused by labor constraints
• Cyber and data risks that come with connected operations

For this topic series—AI in Supply Chain & Procurement—ARC is also a natural “checkpoint” event. It’s where leaders compare approaches across:

• AI forecasting and demand sensing
• Supplier risk monitoring
• Transportation optimization (routing, tendering, appointment scheduling)
• Warehouse automation and workforce orchestration
• Control towers and visibility platforms that feed AI decisioning

If you’ve been building in any of those areas, ARC is asking for what the market actually needs: case studies, practical lessons, and strategic frameworks—delivered in a vendor-neutral way.

What “real change” looks like (and what most teams get wrong)

Answer first: Real change is when AI becomes part of daily decisions—planning, execution, and exception management—with metrics tied to cost, service, and risk.

Most companies get this wrong in one of two ways:

1. They deploy AI only in planning. The model predicts demand, but execution never changes. The warehouse still waves orders the same way. Transportation still tenders the same way. The result: smart forecasts, dumb outcomes.
2. They automate decisions without governance. The model makes recommendations, but nobody can explain why, validate drift, or handle edge cases. The result: short-term gains, long-term fragility.

If you’re considering a speaking proposal, anchor it around the operational “moments that matter,” such as:

• How your team handles late inbound shipments and re-plans allocation
• How you decide expedites vs. substitutions under constraints
• How procurement adjusts sourcing when tariffs or sanctions hit a lane
• How you rebalance inventory when demand shifts across regions mid-quarter

A practical definition you can use on stage

Intelligent operations: systems and teams that sense changes, predict outcomes, and take or recommend actions fast enough to matter—inside planning and execution.

Supply chain resilience: the ability to maintain service and margin when constraints change, by adapting sourcing, inventory, capacity, and fulfillment policies.

These definitions are “stage-ready,” and they keep the conversation grounded.

Three session angles that audiences will actually remember

Answer first: The best talks show a before/after operating model, the data and process changes required, and the measured impact.

The RSS source highlights themes like AI, resilience, geopolitics, and knowledge transfer. Here are three session angles that map to those themes—while staying relevant to transportation and logistics leaders.

1) AI for resilient planning: from forecasts to decision velocity

A strong resilience talk isn’t “we used machine learning.” It’s how quickly the business can respond.

What to cover:

• Where AI sits in your planning stack (demand, supply, inventory, transportation)
• Your approach to scenario planning (what you simulate, how often, who approves)
• How you measure improvement (examples that land well):
  • Forecast error reduction at item-location level
  • Service level during disruption windows
  • Inventory turns without service degradation
  • Expedite spend as a percent of sales

A concrete story format that works:

1. The disruption (port congestion, carrier capacity squeeze, supplier failure)
2. The old process (manual triage, weekly cadence)
3. The new process (AI-driven exception prioritization, daily cadence)
4. The metric change (cost, service, working capital)

2) AI in transportation: optimization that respects real constraints

Transportation optimization fails when it ignores execution realities: appointment windows, dock constraints, driver hours, detention, and customer-specific rules.

A compelling session shows how you handled constraint complexity, for example:

• Using AI to predict ETA risk and trigger preemptive replanning
• Combining optimization with carrier performance scoring to improve tender acceptance
• Automating accessorial detection (detention, layover) with audit logic
• Improving last-mile outcomes with dynamic routing tied to customer preferences

One line I’d use if I were presenting:

If your optimizer doesn’t know your docks are full, it’s not optimizing—it’s guessing.

3) Knowledge transfer: the hidden bottleneck in AI supply chain deployments

The call for speakers explicitly mentions “unlocking knowledge transfer.” That’s not a soft topic—it’s one of the highest ROI enablers of AI.

Why? Because AI projects stall when tribal knowledge stays tribal:

• Planner overrides aren’t captured as data
• Supplier relationship context isn’t codified
• Exceptions are resolved in email, not in systems

A practical talk here might include:

• How you structured a decision log (what was decided, why, by whom)
• How you built playbooks for exceptions (and made them machine-readable)
• How you reduced key-person risk in planning and dispatch

What ARC audiences expect: vendor-neutral, measurable, repeatable

Answer first: ARC audiences want to learn how to replicate outcomes, not how to buy a product.

The submission guidance from the RSS source is clear: real-world case studies, practical lessons, strategic frameworks, and vendor-neutral delivery.

If you’re writing an abstract, build it around these elements:

• Problem statement (one sentence, business-first)
• Constraints (data quality, change management, union rules, system limitations)
• Approach (what you changed in process + data + tech)
• Results (quantified, with timeframe)
• Lessons learned (what you’d change next time)

Metrics that make a proposal credible

You don’t need “perfect” metrics, but you do need real ones. Pick 3–5, such as:

• On-time in-full (OTIF) or on-time delivery
• Order cycle time
• Tender acceptance rate
• Detention hours per load
• Forecast accuracy / WAPE by segment
• Inventory turns / days of supply
• Expedite cost and premium freight frequency
• Supplier OTIF and lead time variability

Even better: include the trade-off you managed.

The best resilience stories include a sacrifice you avoided. “We improved service without bloating inventory” is a stronger narrative than “service improved.”

A speaker-ready outline you can adapt (150–250 word abstract friendly)

Answer first: Use a structure that telegraphs business impact, method, and repeatability in under 250 words.

Here’s a template you can steal and customize:

• Title: “From Exceptions to Outcomes: AI-Driven Logistics Decisions at Scale”
• Abstract (core beats):
  • The operational pain (late inbounds, volatile demand, constrained capacity)
  • The old workflow (manual triage, slow replans, inconsistent decisions)
  • The AI-enabled workflow (prediction + prioritization + recommended actions)
  • How you governed it (human-in-the-loop, drift monitoring, override logging)
  • Outcomes (service, cost, working capital) over a defined timeframe
  • What attendees can replicate in 30/60/90 days

Suggested attendee takeaways (ARC-style)

• A checklist for choosing where AI belongs (planning vs execution vs both)
• A simple governance model that prevents black-box failure
• A data readiness approach focused on operational signals (appointments, dwell, exceptions)
• A measurement framework that ties AI to cost-to-serve and resilience

People also ask: “What should I speak about if I’m not a data scientist?”

Answer first: Speak about the operating model—because that’s where AI wins or dies.

You don’t need to present model architecture to contribute. Some of the highest-value talks come from:

• Operations leaders who redesigned exception workflows
• Procurement leaders who changed supplier segmentation and risk triggers
• Transportation managers who operationalized prediction into dispatch decisions
• Program owners who solved cross-functional data ownership

If you can explain how AI changed day-to-day decisions—and what metrics moved—you’re qualified.

Call to action: bring the work, not the hype

If you’re working on AI forecasting, supplier risk analytics, warehouse automation, routing optimization, or control-tower decisioning, ARC Industry Forum 2026 is a stage where practical results are the currency.

A strong proposal doesn’t need buzzwords. It needs honesty: the messy middle, the change management, the data fights, the moments when the model was wrong—and what you did next.

If you could help one peer avoid a six-month detour on their AI in supply chain & procurement roadmap, what would you tell them from the podium?