AI is now the central focus in TMS discussions among leading vendors, with 60% of Oracle’s customers already deploying AI tools. Experts from Caroz, Manhattan Associates, and Oracle emphasize embedding AI into core business processes—not as add-ons—while stressing data quality, unified systems, and hybrid operational models. Samsung reduced transport costs by 22% using Caroz’s control tower; Manhattan reports 87% of companies doubt their current TMS’s future-readiness; and Oracle’s embedded AI cuts implementation to 3.5 months. Real-world ROI includes ~20% lower logistics costs for AI leaders versus laggards, per McKinsey-Oracle research.
According to www.supplychainmovement.com, artificial intelligence is the dominant theme in virtually every conversation among transport management system (TMS) suppliers — a trend confirmed by experts from Caroz, Manhattan Associates, and Oracle during Supply Chain Media’s Webinar Wednesday ahead of the TMS Demo Day on 15 April 2026.
Strategic Priorities Align Around Resilience and AI
Research by Supply Chain Media identifies greater resilience and agility as the top goals of supply chain transformations. Adoption of artificial intelligence and implementation of modern software solutions like TMS are tied for fifth place — yet these two priorities are deeply interlinked. As Rob Hazekamp, director of logistics solutions at Oracle, states:
“That’s what all the conversations are about.” — Rob Hazekamp, director of logistics solutions at Oracle
Practical AI Adoption Pathways
Martijn Lofvers, moderator of the webinar, advises companies to begin AI adoption with a data lake to unify fragmented IT systems. Mark Vos, principal solution consultant at Manhattan Associates, emphasizes embedding AI directly into business processes:
“Companies are eager to explore the opportunities offered by AI and to discover how they can use this technology to cut costs and streamline processes.” — Mark Vos, principal solution consultant at Manhattan Associates
However, Maurits Jongens, managing director of TMS at Caroz, cautions that a data lake alone does not guarantee data quality:
“A data lake is very welcome, but it does not automatically mean that data quality is up to standard. We see many companies struggling with this, which makes it difficult to deploy AI successfully.” — Maurits Jongens, managing director of TMS at Caroz
Hybrid Control Towers and Real-World Impact
Caroz offers hybrid control towers — where shippers manage day-to-day operations internally while engaging Caroz specialists for rate benchmarking or strategic alternatives. Samsung, a Caroz user, achieved a 22 per cent reduction in transport costs through more efficient routing and smarter carrier allocation. Jongens notes shifting market dynamics: companies pivoted from contract carriers to the spot market pre-pandemic, then reversed course amid capacity shortages — now favoring hybrid models.
Unification Over Integration
Manhattan Associates frames TMS not as a standalone tool but as part of its unified Supply Chain Execution suite, encompassing warehouse, yard, and labour management. Vos rejects siloed thinking:
“We want companies to stop thinking in terms of TMS or WMS, but rather in terms of, for example, incoming and outgoing goods flows.” — Mark Vos, principal solution consultant at Manhattan Associates
Manhattan’s research shows 87 per cent of companies question whether their current TMS can keep pace with evolving transport market demands. Visibility, integration with carriers, and unification across systems are critical — because transport “is not a stand-alone process, but an integral part of supply chain execution.”
Embedded AI Delivers Measurable ROI
At Oracle, AI is built into applications — not bolted on. Hazekamp cites joint research with McKinsey showing AI frontrunners achieve logistics costs ~20 per cent lower than laggards. Around 60 per cent of Oracle’s customers are already using AI in business operations. One application delivers real-time recommendations to transport planners — supporting (not replacing) human decision-making, with automatic implementation options. Centralized best practices enable configuration adaptation via AI, cutting average implementation time to 3.5 months and shortening payback periods.
Source: www.supplychainmovement.com
Compiled from international media by the SCI.AI editorial team.
