Technology disruptors are going to have a huge impact on the transportation industry in the coming years. But they may not be the disruptors you’re looking for. In 2020, self-driving trucks and drone deliveries were all over the industry press and were thought to be a coming sea change in efficiency and productivity. However, a recent survey by Princeton Consultants showed that only 40% of respondents believed drones would have a moderate/large impact in the next seven years, and an even lower 34% believed self-driving trucks would have a moderate/large impact. So, what changes are going to be the drivers? The Internet of Things (IoT) led the pack, with 74% of respondents saying it would have a moderate/large impact in the coming years. Artificial intelligence/machine learning came in a close second at 67%.

As it relates to transportation, IoT means three things: sensors, sensors, and sensors. Visibility, early warning, and real-time intervention are the name of the game here. Milestone-based tracking has been the default since the early 1990s. In recent years, both national carriers launched consignee-managed delivery management tools and shipper-managed rerouting tools for pharma and other special use cases. For now, these carrier-provided tools come with pretty steep costs.

As the technology improves and end-to-end visibility becomes ubiquitous, costs should come down. Carriers like to inflate prices for value-added services, pointing to their fees as pass-throughs. But over time, the truth is these charges become margin drivers. However, in this case, third-party providers are probably the solution to the price problem. There are three key elements to this type of solution:

  • Implementation of low-cost, durable sensors
  • Integration with a machine learning-capable backbone
  • Ability to take corrective action at scale

Companies like project44 and Fourkites already offer end-to-end visibility solutions. As visibility becomes more contemporary, machine learning will become more important so that exceptions become predictive rather than reflective, and mitigation can become proactive rather than reactive. The question remains how deeply parcel carriers will allow third parties to integrate into their system for redirects and service changes. Historically, UPS and FedEx have been antagonistic toward third-party service providers — mainly when those providers compete directly with the services they provide. In this case, however, both UPS and FedEx have existing relationships with some of these companies, so there may yet be a path for deeper integrations.

This brings us to machine learning. In my view, this will be the main game changer in the next three to four years. Machine learning is the necessary unlock to turning real-time visibility into actionable intervention, allowing the real-world experience to dictate what variables lead to late packages, flagging those packages, and instituting corrective action. Moreover, machine learning tells us what we don’t know. Everyone has their key performance indicators that they track on a regular basis, metrics based on hard-won lessons that tell you what has bitten you in the past. But what about those things that haven’t bitten you yet? Machine learning can intake data and, based on patterns indiscernible or apparently irrelevant to a human analyst, learn what is normal for you and flag the exceptions. New packaging resulting in a greater percentage of DIM’d packages? Learn about it in a week instead of a quarter. Fraud on your account? We have seen machine learning pick up on this almost instantly. In fact, there’s practically no limit to what ML might pick up, as it’s not looking at a list of measurable variables. Instead, it’s looking for exceptions to the norm.


  • An unexpected appearance of “automatic warranty” charges for a customer that, historically, had no prior charges for this accessorial. After consulting with the carrier, it turned out that during the uprate, the ratings team incorrectly applied this charge. As a result of this finding, the carrier issued the shipper a credit for $138,443.
  • A customer experienced a per diem charge that was not only its highest historical charge for this accessorial to date, but it was also sent by the carrier with no additional documentation. Meaning, we had no visibility as to how this charge was calculated. After bringing this to the customer’s attention, the entire invoice was short paid and disputed for a total savings of $75,141.
  • Machine learning flagged disproportionate costs for shipments coming out of a Texas location. It turned out they’d had a team member come from Europe to that delivery center in Texas. In Europe, they use a comma the way we use a decimal point in the US. But the software in Texas wasn’t treating the comma as a decimal point, so instead of a five-pound package it was interpreting the weight as 500 pounds. After consulting with the customer, they advised us that this charge was incorrect. As a result, the charge was disputed, and the customer received a $11,562 credit from the carrier.
  • A customer received multiple invoices from a carrier for a new and anomalous sender address. After bringing this to the customer’s attention, it was discovered that the carrier had sent incorrect EDI references, and these invoices were sent in error. As a result, the carrier voided these invoices, totaling $649,134.

So, where can shippers find these magical benefits? I’ve already mentioned a few vendors that are providing real-time visibility. Some audit and visibility providers are offering machine learning integrations and services. And other innovators will spring up as the market further defines the need. It is an exciting time to be in the transportation space. Like the technology we use, we as practitioners must evolve our processes, strategies, and even roles to keep pace with enabling innovations. Peak shipper-side efficiency may be an unreachable goal, but we’re closer than we’ve ever been.

Joe Wilkinson is VP, Professional Services (Transportation Consulting) at Intelligent Audit. He can be reached at

This article originally appeared in the May/June, 2023 issue of PARCEL.