The Human Algorithm: ArcBest’s AI Journey to Smarter, Scalable LTL Operations

Optym LTL Convergence 2025 Session with Dennis Anderson, Chief Innovation Officer at ArcBest

Dennis Anderson of ArcBest shared what happens when a century-old company decides to reinvent itself with AI - without losing human touch. This wasn’t about buzzwords or lab experiments. It was a grounded, results-driven case study on how ArcBest used artificial intelligence to simplify the complexity of LTL, scale faster, and empower its people.

A Supply Chain That Never Stands Still

“The supply chain is more dynamic than ever,” Anderson said. “Customer expectations are rising fast - for speed, visibility, and transparency - and legacy systems can’t keep up.”

Those challenges became a turning point for ArcBest’s LTL carrier, ABF Freight. The company saw opportunities where others saw disruptions. By modernizing its LTL network with AI and data, ABF Freight positioned itself to handle rising costs and customer expectations while maximizing resource utilization. At the center of that transformation is City Route Optimization (CRO), an AI system built in-house to make every mile, stop, and driver count.

Turning Local Know-How into Machine Intelligence

For decades, route planning at ABF was a process that relied heavily on manual intervention and local knowledge. A supervisor, armed with tools that only allowed for planning at the zip code level could spend four plus hours building the next day’s delivery plan depending on the size of the operation. CRO changed that.

By combining human expertise with machine learning, they turned local knowledge into scalable intelligence. The planning time was cut in half, while the AI’s plans began consistently outperforming manual ones.

“We learned early that it’s not about man versus machine,” Anderson said. “It’s man with machine.”

Building AI Around Real Work

The ArcBest and ABF Freight teams rolled out CRO in three practical phases, each designed with operators in mind:

Phase 1: Deployed static route planning across all 240 service centers, embedding real-time feedback from planners and drivers.
Phase 2: Introduced dynamic route adjustments, allowing the system to react to variability in day-to-day operations, new appointments, and changing customer requirements.
Phase 3: Focused on pickup efficiency, predicting when and where to position drivers throughout the day for an improved customer experience.

At every step, planners were invited into the process. Through A/B testing, they compared their manual routes to AI-generated ones. The AI models didn’t replace their judgment, they amplified it.

The Numbers Behind the Change

ABF’s approach paid off in measurable results:

$13 million in annual savings from Phase 1 alone.
17% reduction in cartage agent usage, keeping more deliveries in-house.
1.5% productivity improvement across all service centers.
3.3 million miles removed from the network.
4.6 million metric tons of CO₂ gases avoided through more efficient routing.
Planning time cut by more than 50%.

Those numbers represent a shift to smarter, faster and more sustainable operations.

Making Technology Feel Like Help, Not Change

Anderson admits the rollout wasn’t easy. “We discovered early that if we didn’t frame AI as help, it felt like competition,” he said. By positioning technology as a partner to experienced planners, ArcBest built trust. The CRO system was refined alongside the people who used it every day, ensuring it worked the way real operations worked.

The result was rapid adoption and continuous refinement. Every service center became a testing ground, feeding data back into the model to make it stronger and more accurate over time.

Balancing Local Realities with Network Goals

Each of ABF Freight’s 240 service centers has its own terrain, traffic, and customer mix. What works in Baltimore may not work in Dallas. CRO was built to flex with that reality. It balances local efficiency with overall network performance, finding the sweet spot where both can win.

That combination of local precision and network-wide visibility has become a competitive edge, a model for how legacy carriers can modernize without losing operational intuition.