skip to main text

Tin tức toàn cầu Logistics automation will work with humans, not replace them

Ngày đăng kíMAR 02, 2023

Eric Johnson, Senior Technology EditorFeb 17, 2023, 8:00 AM EST
Articles reproduced by permission of Journal of Commerce.

Eric Johnson, Senior Technology Editor
Feb 17, 2023, 8:00 AM EST
Articles reproduced by permission of Journal of Commerce.

Logistics automation will work with humans, not replace them Logistics technology providers believe artificial intelligence tools will eventually make supply chain data more easily searchable and usable. Photo credit: jittawit21 /

Automation is slowly transforming the logistics industry, not in the ominous job-killing sense, but in the already widespread adoption of robotic process automation (RPA) for repetitive tasks and data entry, according to a variety of technology experts that spoke with the Journal of Commerce.

Marc Held, CEO of automated freight insurance software provider Fishtail, said the attention paid to automated machines that can perform physical freight movement tasks normally handled by humans might be obscuring a more important evolution that has been underway in the background for years.

“With every generation of technology — from spreadsheets to ERPs [enterprise resource planning systems] to RPA — it becomes more and more pronounced that there are some things that computers are really good at, and some things that humans are really good at,” he said. “It’s pretty clear that people won’t be replaced for quite some time, but we are starting to see some pretty subtle ways that businesses are transforming to take advantage of these new tools — doing more with less, doing completely new things, and shaving off cost.”

Logistics software providers and investors harbor dreams of autonomous environments in which global supply chains composed of connected systems regulate themselves with minimal input from people.

But between now and that future state, RPA is already taking root in areas that are hiding in plain sight, said Liz Ward, head of partnerships at ZEBOX, the CMA CGM-backed technology incubator.

Similar to the prerequisite courses a student must take before moving on to more advanced classes, RPA is a foundational element of artificial intelligence (AI) and machine learning models. Using RPA to automate data entry, information extraction, or responses to customer service enquiries, for example, can lead to automation of more advanced processes such as procurement or routing decision-making.

Ward cited increased use of optical character recognition (OCR) tools for back-office functions, including data entry, freight bill audit, contract reconciliation, payments, and quoting, as an area in which RPA is already changing supply chain processes for the better. She also noted an emerging category she dubbed “communication aggregators,” platforms that automate the collection of messages across different software channels and automates the prioritization of tasks associated with those messages.

“Automation is thriving in areas where there are fewer variables, constraints, nodes, modes, and people involved,” she said. “All the reasons why things like freight brokerage and macro demand planning platforms are kind of automation pipedreams. Too much stuff can go wrong, making it a back-end engineer’s worst nightmare.”

In that sense, Held said automation in the near term will be much more a matter of “human and machine” than “human versus machine.”

“Years ago, it would’ve been unthinkable to have technology that tells you how you should be buying and selling freight based on market conditions and what your company specializes in, but now it’s pretty much state of the art,” he said. “It would’ve been unthinkable to not have teams of people manually keying in the information that lives in bills of lading, invoices, or other trade documents.” Focus on forwarding Much of this activity is taking place in the forwarding software space, as third-party logistics providers look to technology to increase employee efficiency and expedite customer service processes such as rate queries and invoicing.

For example, RPA can generate human-like text that can be used to automate responses to customer inquiries, improve communication with suppliers, and provide real-time updates on shipments, noted King Alandy Dy, CEO of freight automation specialist Expedock, which caters to forwarders.

Alandy Dy said this use of RPA in discrete processes will continue to increase as developers of automation tools purpose-built for the logistics industry intersect with more general tools like the chatbot program ChatGPT.

“We have calibrated an AI, chatbot-like interface that can inspect a business’s supply chain data and answer queries in human-understandable responses,” he said. “Personally, we believe that this will be one of the ways that businesses will interact with their supply chain data in the future.”

Alandy Dy expects such capability to eventually be applied to other core supply chain systems in ways that will improve forecasting and in-the-moment decision-making. ChatGPT might also be able to provide insights from unstructured data sources such as emails, customer reviews, and social media posts — long seen as a “holy grail” in supply chain automation.

Forwarder Geodis, for example, worked with an RPA specialist last year to create automated responses to common questions in requests for information from prospective customers, decreasing the amount of time its employees spend replying to those inquiries.

“There’s a wave of these new AI models that can drastically improve existing supply chain processes, and the innovation is just beginning,” Alandy Dy said. ‘Self-explanatory’ value But for shippers, the crucial next step in the automation journey is linking enterprise ERP systems with external visibility platforms, allowing for better alignment of order generation and transportation functions, said Graham Parker, CEO of technology provider Gravity Supply Chain Solutions.

“Some shippers and logistics service providers are doing this already, but nowhere near enough,” Parker said. “Automation plays a role not just in the visibility but also in the lifecycle, looking at demand versus production. Demand and order management visibility with intuitive, data-driven actionable insights go hand in hand.

“Shippers are great at digitizing the last 100 yards and online shopping, but the backbone of everything is the supply chain, and this is where there needs to be greater adoption of technology and trust in the data,” he added. “These platforms exist.”

Parker said shipper investment in supply chain automation should have a singular focus: to make “data keying and spreadsheet crunching a thing of the past.”

Once those elements are automated, shippers can track key performance indicators (KPIs) using live data and assess their logistics service providers at any time, rather than in quarterly or annual reviews, he said. Automation of data between systems also plays a pivotal role in allowing AI tools to “identify risk as it becomes apparent versus plan, and often ahead of time,” he added.

The benefits of such an approach are “self-explanatory” in terms of return on investment, Parker said, adding that automation should enable shippers to reduce excess inventory levels and avoid errors from re-keying data.

Parker, however, said companies should be focused more on quickening adoption of automation, rather than granular benefits, because the longer-term benefits come from “fully connected automated data throughout the entire business.”

Fishtail’s Held said adoption of automation among shippers and logistics providers has mostly been limited by a reluctance to “fix” processes that aren’t completely broken, but he believes the industry may be readying to change in ways that will allow automation to flourish.

“What I’m most excited about in the coming years is that we’re starting to finally see companies and technologies integrate in a more coherent manner,” Held said. “I’d argue that the main limiting factor for true automation isn’t really the technology but the fact that the same piece of data may represent completely different things to completely different parties. A container for one company may be seen as a bunch of orders for another company, which may be seen as one or two invoices from that second company’s customer.”
· Contact Eric Johnson at and follow him on Twitter: @LogTechEric.