Logistics have always been data driven. Historically, logistics data can be found on cuneiform tablets, pieces of parchment, and paper. In the Digital Age, paper trails need to give way to cyber records. Most supply chain professionals understand this, which is why the interest in blockchain technology is growing. Blockchain will certainly help logistics operations in a number of ways; but, digital data already available can also help. Shannon Vaillancourt, President of RateLinx explains, “The right data can create the intelligence required to know where to begin to solve many common problems. These can include freight costs that are too high, transit times that are too long or not charging enough for freight.”[1] Vaillancourt understands the “intelligence” created from the data is the result of analysis. There is a symbiotic relationship between data and analytics that creates business value.
Data and Cognitive Computing
Some of today’s datasets are huge. Manual analysis of the data contained in these sets is simply impractical in today’s fast-moving business environment. Complexity adds another layer of difficulty to the logistics picture. Boston Consulting Group (BCG) analysts assert, “Logistics networks have never been more complex. The increasing prevalence of operations that are global — with growing numbers of production sites and clients that could be anywhere on earth — has introduced challenges that didn’t exist to the same degree at an earlier time.”[2] Fortunately, cognitive technologies are maturing at just the right moment to help. Christine Taylor observes, “Artificial intelligence and logistics is the perfect union for businesses seeking a competitive edge — and a necessary combination for companies hoping to compete in the future.”[3] Vaillancourt explains it takes at least four different datasets to gain a complete logistics picture. He writes:
“A shipper needs four datasets to be integrated, cleansed, and standardized: 1) shipment data; 2) freight invoice data; 3) track & trace data; 4) order & item level data. The shipment data shows how the freight was prepared (weight, pieces, dimensions, and accessorials). The freight invoice data verifies what really happened (Were measurements correct? Were other accessorials required?). The track & trace data shows if any exceptions occurred once the freight left your warehouse (weather delays, mechanical problems, etc.). The order and item information shows what was shipped and to whom (Was it a back order? Is this a customer with a specific routing guide? Does this item require special handling? etc.). By having all four datasets, you now have the proper insight into why, and it helps you understand what you can do (if anything) better.”
Vaillancourt seems to assume most companies have these datasets available in an integrated logistics platform. BCG analysts report, “Few have a holistic view of their logistics network or of the associated costs.” They explain:
“Such a view considers all supplier and customer flows simultaneously. It tracks not only transportation and handling but also warehousing and inventory working-capital costs, specifically by end customer. Without such a picture, companies might miss opportunities to capture greater value from their network. The incomplete view that most companies have of their transportation costs is the result of a siloed organization structure that has each part of the business focusing on a single aspect of logistics.”
A cognitive platform can help integrate siloed data so a holistic picture can be obtained; however, a cognitive platform can do much more than that.
Logistics and Cognitive Computing
Ubiquitous sensors connected by the Internet of Things (IoT) to a cognitive platform can provide real-time logistics intelligence. Jim O’Donnell (@jimodonnelltt) notes, “The availability of inexpensive sensors and internet of things connectivity has made supply chain visibility easier for manufacturers in the past few years. In fact, it’s possible to know exactly where your goods are at any time and, in many cases, what condition they’re in.”[4] He adds, “It’s great to know where your goods are at the moment, but wouldn’t it be better to know exactly when you’re going to get them? That’s becoming more likely as internet of things and sensor data increasingly combine with artificial intelligence, machine learning and other next-generation data analytics tools that provide predictive logistics to help manufacturers go far beyond visibility of supply chains.”
The staff at Material Handling & Logistics (MH&L) suggests cognitive technologies will play a role in almost every area of logistics. “AI in logistics,” the staff writes, “will include intelligent logistics, predictive operations, back-office automation and new customer experience models.”[5] Citing a study conducted by IBM and DHL, the staff concludes:
“AI has the potential to significantly augment human capabilities. While AI is already ubiquitous in the consumer realm, as demonstrated by the rapid growth of voice assistant applications, AI technologies are maturing at a great pace, allowing for additional applications for the logistics industry. These can, for instance, help logistics providers enrich customer experiences through conversational engagement. With the help of AI, the logistics industry will shift its operating model from reactive actions to a proactive and predictive paradigm, which will generate better insights at favorable costs in back office, operational and customer-facing activities. For instance, AI technologies can use advanced image recognition to track condition of shipments and assets, bring end-to-end autonomy to transportation, or predict fluctuations in global shipment volumes before they occur. Clearly, AI augments human capabilities but also eliminates routine work, which will shift the focus of logistics workforces to more meaningful and value-added work.”
Taylor reports DHL believes two of the biggest usage trends in AI and logistics are anticipatory logistics and self-learning systems. According to Taylor, “Anticipatory logistics are based on predictive algorithms running on big data. The practice allows logistics professionals to improve efficiency and quality by predicting demand before a consumer places an order. … Anticipatory logistics also serves supply chain risk management. AI predicts maintenance needs and potential risks, similar to transportation and disruption management predictions. Manufacturing and transportation industries use AI to predict factory and vehicle maintenance. In this case, predictive maintenance is based on sensor data gathered from smart machines and vehicles.” Turning to self-learning systems, Taylor writes, “The logistics sector has been slow to adopt machine learning, but forward-thinking logistics companies are adopting self-learning systems. Self-learning logistics systems improve their algorithms as they get more data over time. … A newer development in self-learning systems is intelligent warehouses. These systems recognize repeated trends and incidents, analyze the repeated data, connect the data to specific entities such as orders or customers, and launch pre-pack instructions. Another common example is AI and robotics that check on stock levels to reorder and restock as needed. Over time, self-learning enables the system to improve its algorithms for even more accurate responses.”
Summary
The MH&L staff concludes, “AI stands to transform the logistics industry into a proactive, predictive, automated and personalized branch.” Taylor adds, “Artificial intelligence and logistics may not be The Jetsons yet. But the combination is fundamentally altering supply chain management across the world.”
Footnotes
[1] Shannon Vaillancourt, “The Answer to Logistics Problems is Always in the Data,” Inbound Logistics, 25 April 2017.
[2] Andrew Loh , Simon-Pierre Monette , Andres Garro , Dustin Burke , Andreanne Leduc , and Nicholas Malizia, “Leveraging Big Data to Manage Logistics,” Boston Consulting Group, 16 February 2016.
[3] Christine Taylor, “Artificial Intelligence and Logistics is Transforming Business,” Datamation, 25 October 2017.
[4] Jim O’Donnell, “Using predictive logistics to improve and expand supply chain visibility,” TechTarget, October 2017.
[5] Staff, “Artificial Intelligence to Thrive in Logistics,” Material Handling & Logistics, 25 April 2018.
