If logistics companies have been sleeping on artificial intelligence and its potential to transform the industry, it’s time for them to wake up and smell the algorithms, according to the new “Artificial Intelligence in Logistics” report by DHL and IBM.
There’s good reason for the logistics industry to sit up and pay attention to AI. First, significant improvements in computer processing speed and power mean AI is getting much faster—and much more capable. Coupled with AI systems’ greater access to big data and better algorithms that can handle more complex applications, artificial intelligence is becoming more powerful by the day.
Plus, AI is following the typical tech adoption cycle: a new innovation is a hit with consumers, then makes the transition into the commercial enterprise and finally makes its way into industrial applications. As consumers, we’re so used to some AI applications—like Amazon recommendations, for example—that we hardly even think of them as anything out of the norm. Chatbots have been among the most talked-about commercial AI applications, especially in retail. So it’s just a matter of time before industrial groups like the logistics industry gets on board with AI, too.
What’s more, all of the major titans of tech are adopting AI-first strategies, from Amazon, Google and Microsoft to Alibaba and Baidu. Investment in AI is heating up even as adoption remains low. In 2017 100 AI startups raised $11.7 billion in 367 deals.
Supply chains already are a rich source of structured and unstructured data; with the ongoing transformation to digital supply networks, incorporating artificial intelligence is the natural next step.
Related Stories
AI affords a level of efficiency and optimization not possible through human brainpower alone. Though much has been made of AI’s threat to jobs, many experts believe that artificial intelligence can eliminate many of redundant tasks and free up professionals to focus on more high-value activities. CAD and engineering software provider Autodesk used the IBM Watson Conversation platform to develop a virtual agent that answers 40 different simple queries, executing about 30,000 interactions monthly, and reducing response time by 99 percent—from 1.5 days to five minutes.
Another AI application is in an “expert assist” capacity that would reduce the amount of time professionals spend searching unwieldy databases and corporate intranets. AI could use specific technologies to reduce this search time by 50 percent.
There are a number of AI use cases focused specifically on the logistics industry. Using AI natural language processing technologies to detect financial anomalies could reduce the burden on logistics accounting teams that process millions of invoices annually from vendors, providers and partners.
A “cognitive” approach that uses AI for customs brokerage could streamline the process significantly. An AI system could be “fed” all of the legislative materials, regulatory documents, customs brokerage SME knowledge, and relevant handbooks to figure out how to automate the process of customs declarations. Human intervention would help whenever the AI system encounters an exception.
Remember the fidget spinner fad last year? Seemingly out of nowhere, the toy became a must-have item, selling 50 million units over several months and accounted for 20 percent of all retail toy sales during this period. Though this type of product normally would enter the U.S. as ocean freight, retailers desperate to capture sales turned to air freight and express networks to satisfy demand and keep inventories stocked, inundating shippers with volume—and perfectly illustrating why the industry needs predictive demand and capacity planning.
Other innovations like autonomous guided vehicles (AGVs) that could replace forklifts and wheeled totes to transport goods safely are already use but will become more widely adopted. Logistics companies should also take note of a proof-of-concept that IBM Watson undertook to demonstrate AI and computer vision for visual asset inspection. The idea involved cameras installed on bridges that photographed cargo train wagons passing beneath. Watson analyzed the photos over time to determine any damage to the train cars and what type it was, and assign it for repairs. In a short period of time, Watson achieved a visual accuracy rate of 90 percent.
Yet other AI developments are putting voice technology into hyperdrive. AVRL’s voice technology allows logistics workers to interact with their IT systems conversationally, using colloquial or informal speech as if they were speaking with another human. It can accurately connect the mention of a product name to its corresponding product information in ERP, WMS and TMS systems. “The ability to automate input, store, and retrieve information via conversational voice interaction removes time and complexity from many warehouse tasks that require manual input or lookup of information,” the report noted.
Predictive capabilities also apply to network management, because while air cargo accounts for just 1 percent of global trade by tonnage it’s also 35 percent by value, according to the report. On time and in-full shipments are especially critical for air freight.
DHL saw the need to better address this issue and built a machine learning-based tool to predict air freight transit time delays and thus help to proactively mitigate problems before they start.
The machine learning model analyzes 58 data parameters to predict if the average daily transit time for any given lane is expected to rise or fall as far as a week in advance. It can also detect the top factors influencing shipment delays, such as day of departure or the airline’s on-time performance.
“This can help air freight forwarders plan ahead by removing subjective guesswork around when or with which airline their shipments should fly,” the report noted.