CURRENT STATE

AI today: early progress, lingering barriers

If AI adoption in transportation management were a marathon, most industry stakeholders would still be at the starting line. While some have taken their first steps, the real journey is just beginning.

How will AI transform operations? What does the AI-powered future of transportation management systems (TMS) look like? How can businesses prepare to seize the moment? Throughout our executive surveys and interviews one thing was clear: everyone is eager to implement, but uncertain how.

Below, we look at the industry's current attitudes and perceptions about the technology in all its forms.

SHIPPERS

Where in your transport management do you use AI-enabled tools?

According to our survey, nearly half of shippers (44%) are using some form of AI in transportation planning and optimization, and many are experimenting with freight procurement (37%) and real-time visibility (32%) too.

Check full data →

Transportation planning & optimisation

Freight procurement

Real-time visibility

Pricing and lane optimization

Real-time tracking and ETA management

Driver scheduling and route planning

CARRIERS & LSPs

Where in your transport management do you use AI-enabled tools?

Carriers are doing the same in areas like pricing and lane optimization (42%) and real-time tracking (39%).

Check full data →

AI-powered TMS adoption

Check full data →

While 36% of shippers reported having moderate or basic AI capabilities in their TMS, about 25% said they’re not using AI at all, and about 18% don’t even have a TMS.

Among those that do, only a small fraction (1%) report advanced capabilities such as autonomous decision-making.

Most still rely on basic, rules-based automation.

Even for those investing in AI, progress is slowed by a familiar obstacle: data quality.

More than half of both shippers and carriers cite poor or inconsistent data as the biggest barrier to success. Integration issues compound the problem, as many struggle to connect internal systems and external partners. AI models can’t make accurate predictions when the data feeding them is incomplete, delayed or siloed.