A wave of AI-driven change is approaching transportation management, with more than half of companies expecting to deploy autonomous agents by 2030, according to Manhattan Associates. But despite the optimism, many are still grappling with foundational gaps that could stall adoption.
Vision Outpaces Readiness in the Push for Agentic AI
Nearly two-thirds of companies now expect fully autonomous AI agents to drive transportation decision-making within five years, according to new data from Manhattan Associates and Vanson Bourne. These agents, designed to execute complex tasks with minimal human input, represent a clear shift from AI as a passive tool to AI as an independent operator. Yet only 37% of firms report deep AI or machine learning integration in their current TMS platforms, signaling a wide readiness gap.
Even among companies that feel confident about agentic AI, 99% admit they either face or expect to face major obstacles. Talent shortages top the list, with 49% citing lack of AI expertise, followed closely by integration challenges (44%) and poor data quality (44%). And while 60% have connected their TMS with broader S&OP systems and 56% have deployed predictive tools, only a minority are using capabilities like real-time demand sensing (35%) or automated booking (36%), systems that will be critical for autonomous agents to function effectively.
Visibility, Volatility, and the Cost of Inaction
The desire for more resilient, efficient transportation networks is undeniable. Nearly 80% of respondents see TMS as a strategic priority today, and that number rises to 86% when looking ahead to 2030. But execution lags expectations. Roughly half of all organizations still struggle with dynamic rerouting and labor scheduling, core capabilities that must be in place before handing control to autonomous systems.
Sustainability remains another stumbling block. While 69% cite sustainability as a top-down mandate, just 34% have embedded it into operations, and even fewer reflect it in procurement or fuel decisions. This limits the ability of AI to optimize for both cost and carbon, a growing pressure as regulations like the EU’s CSRD begin reshaping logistics reporting requirements.
Even small inefficiencies are proving costly: 48% of firms say more than 10% of their logistics spend is lost to disruptions and errors. Yet confidence remains high that AI-led advances in planning and forecasting will cut freight costs by at least 5% over the next five years. The key question is whether organizations can build the technical and operational foundations fast enough to realize those savings.
Don’t Let Strategic Intent Drift Into Technical Debt
The growing enthusiasm around autonomous TMS capabilities risks obscuring a hard truth: without closing today’s visibility, integration, and talent gaps, agentic AI may add complexity instead of solving it. One overlooked factor is how uneven data maturity across business units can create decision friction, undermining the very autonomy these systems aim to deliver. Companies may need to think less about front-end deployment and more about standardizing the foundational inputs that give AI room to work.
