How AI Can ‘Remove All Human Touchpoints’

Artificial intelligence is likely to disrupt the transportation industry, changing the way supply chains are managed and reducing the number of jobs performed by humans, analysts and industry insiders say.

Industry executives say sidewalk robots, self-driving trucks, customer-service bots and even generative AI that can predict disruptions and explain why sales forecasts are off are on the rise.

Morgan Stanley analysts led by Ravi Shankar said last month in a research note that “AI could completely (or almost) eliminate human touchpoints in supply chains, including ‘back-office’ operations.” .

“The freight transportation sector is in the midst of a generational shift driven by disruptive technologies such as autonomous driving, EVs, blockchain and drones. AI is the latest among these potentially transformative technologies. and probably the most powerful ever,” the analysts added.

For example, the bank expects hundreds of self-driving trucks to come into service in the U.S. in 2024, cutting costs per mile by 25% to 30% and ultimately eliminating the need for drivers entirely. (the timescale is “more than 3 years”).

Supply chains are often long and multifaceted. A company may source from manufacturers around the world, ship parts to a central assembly plant, and then ship the goods to customers around the world.

The already complex process of producing and transporting goods has been disrupted by the COVID-19 pandemic and the Russo-Ukrainian war, leading to shortages of components such as computer chips and changes in delivery routes. This complexity means that companies are often unaware of what happens to their products from one end of the process to the other.

“This is where AI (and machine learning) comes in. By predicting what will go wrong in a fluid traffic network … before it happens, AI/ML systems can … avoid chaos scenarios entirely. There is even potential,” added the Morgan Stanley analyst.

It’s a topic picked up by analysts at investment firm Jefferies, who have made multiple predictions about the impact of generative AI on transportation and logistics. This includes forecasting demand, predicting when trucks need maintenance, planning optimal transportation routes, and tracking shipments in real time.

Analysts at the firm, led by Stephanie Moore, said, “Trucker shortages, the polar vortex shutting down interstate commerce, and shortages of baby formula on grocery store shelves are just some of the challenges facing generative AI in the trucking and logistics sector. The introduction will be a distant memory.” said in a research note published June 6.

Generative AI will be a big part of shipping giant Maersk’s business, said Navneet Kapoor, the company’s chief technology and information officer.

“AI and machine learning have been around for a very long time…over the years they have progressed from interesting research projects to more ‘real world’ projects within companies…and now generating With the advent of AI … we have a real transformational opportunity to make AI mainstream,” Kapoor told CNBC by phone.

Maersk has been using AI for several years and is now “actively exploring” ways to integrate AI into its business processes and functions at scale, Kapoor said. One of the ways he’s already used is to help customers plan better.

“We use AI to build so-called predictive cargo arrival models to improve scheduled reliability for our customers. Reliability is critical even after the pandemic. Chains, inventories can be better planned and costs can be reduced.”It’s gone down,” Kapoor said.

Maersk also wants to use AI to recommend solutions when transportation routes are congested, such as whether goods should be flown or stored. And Kapoor said the company wants to use a form of generative AI known as large-scale language models. The model should learn how to recognize, summarize and generate text and other types of content from vast amounts of data and use it to better understand the sales process.

“We have a complete view of every transaction a customer has made in the past year and can get to the root cause. [for example] You can lose deals in certain business areas,” Kapoor said.

And what about potential unemployment?

“What I mean by generative AI is [a] There would be a once-in-a-lifetime turmoil…so jobs would be lost in more traditional situations, but new jobs would be created, as with all previous technological disruptions. I believe,” Kapoor added. Roles such as Prompt Engineer (who trains AI to give better responses) may be in high demand.

One of the threats identified by Morgan Stanley comes from “high-tech digital entrants” to the industry, which analysts say is a double-edged sword for transportation companies: AI could help them become more efficient While it may help, it may also reduce the need for tertiary services – party logistics companies that arrange packing, storage and shipping.

Through its Maersk Growth Ventures division, Maersk has invested in AI start-ups, including self-driving electric truck maker Einride. Pactum is a company that automates deals. 7bridges is an AI platform that helps businesses locate inventory and predict delays.

“We are watching [data startups] This will undoubtedly enable and accelerate our transformation, but we are also cautious. I don’t want to be caught napping on this issue… Data-related startups can: [an] We are intermediaries between our customers and we must always be ahead of the curve, but also learn from them,” Kapoor said.

Igor Licaro, president and chief operating officer of o9 Solutions, a software company that helps companies unify and analyze data, said the Knowledge Assistant could help with another problem: over-ordering or under-ordering goods. This is often due to lack of communication between internal teams, as sales and supply chain management personnel place separate orders, he said.

“This is a sub-optimal result because sales [teams] It’s a waste of money that they may be investing in promoting products that their supply chain restricts,” Ricaro told CNBC by phone.

“We are seeing a world where, hopefully, we will all have what we call knowledge assistants powered by AI and large language models. We will be able to give you insight into why we provided the For example, less than what we ordered.

Answering these questions typically requires input from sales, marketing, supply chain, and procurement teams, but generative AI can potentially examine large datasets to provide answers.

It may also require fewer resources for an integrated business planning team that oversees long-term goals, revenue forecasts and demand forecasts for specific products.

“A planning function that currently has 1,000 people could probably be reduced to 100 or less,” Ricaro says.

— CNBC’s Cheyenne Devon and Jonathan Bunyan contributed to this report.