Sign up for our newsletter and get the latest big data news and analysis.

Data: The Defining Language in Logistics Technologies

In this special guest feature, Marc Meyer, CCO at Transmetrics, discusses AI and just-in-time shipping: how tech models offer a competitive advantage for logistic businesses. Marc is a hands-on commercial strategist with a proven ability to translate business strategies into objectives and action. He has experience in working with startups, upscaling, brand building, lean and agile scrum, and Design Thinking. Marc is a leader in building and implementing an integrated marketing and sales strategies, opening new markets, focusing on brand awareness, client-centric marketing, lead generation, and strategic deal marketing.

When learning a new language, immersion is key. This is because one must truly distinguish the ins and outs of how words are used—not to mention that in linguistics, there are always caveats. Data is a language all of its own. Artificial Intelligence (AI) can help to translate the language of data from a raw form of knowledge to information that contains usable insights and steers action. As codified insider knowledge, AI has the power to transform the efficiency processes of businesses across industries.

This applies especially in logistics technologies where data is the cornerstone of all operations. Emerging AI innovations are a driving factor in defining the usability of data and will be pivotal to the supply chain and all adjacent processes—from dispatch to last-mile. One example is so-called Just-In-Time shipping, named after the famous production method first developed by Toyota. Let’s look at the ways AI, through the language of data, will launch Just-In-Time shipping into the future.

Insight Translated by NLP and ML

It was predicted that AI would increase productivity in logistics by more than 40% by 2035. This is significant because, as with all things in logistics, productivity is not only vital but make-or-break for efficiency metrics for companies.  

Shipping and freight, if anything in this day and age, is international. Globalization has drastically transformed the market and demand is no longer local. It is quite the opposite—shipping companies are often based on one corner of the map, and the products they are moving from point A and B on two other corners. This makes diverse language understanding extremely important. With Natural Language Processing (NLP), AI can derive operational information across multiple languages, capturing the context of all data and pouring all acumen into a central and locally-informed database. 

Equipped with knowledge, AI through Machine Learning (ML) can apply this all-encompassing data to predictive analytics and demand forecasting to better inform physical assets in different areas. A custom-designed AI algorithm can be implemented into linehaul and network investment planning, strategizing more precisely for various regions. This creates insights to help shipping companies anticipate changes in different locations ahead of the demand curve.

Just-In-Time (JIT) shipping is then bolstered because inventory management strategy is better informed on a local level. Many transportation manufacturing companies have their own supply chain divisions—and with better local and regional understanding, the more detailed and accurate the bigger picture is painted. In this way, AI allows for more widespread comprehension across segmented operations, helping to predict demand for each unit within an international or multi-faceted company. This informs companies on how much inventory to have on the ground in a particular location at a specific time—which is especially important when it comes to shipping companies who have inventories of raw materials or perishable goods. The success of this process depends on one thing first and foremost, precise and informed data.

JIT in Logistics Technologies

In a recent study, the extreme value of accurate data was showcased. In the study, a company in question didn’t have direct statistics of the utilization (load factors) of their groupage linehauls because they were only drawing data from a few spot-checks in their linehaul operations. Through an initial analysis, it was shown that less than 20% of the company’s data was actually usable, which was significantly impacting their operations. 

It was found that by extracting information from the Transportation Management System (TMS), the quality of data from back-end procedures has the potential to be improved. Employing an AI-driven data quality framework—using NLP, ML, and computer optimization—a customized algorithm is automated. This algorithm can then be used continuously on a daily basis to help cleanse data on the fly. What is more, it doesn’t affect a company’s existing processes and can run in tandem with established operations.

Within two months, the algorithm as applied to this particular company’s operations reached a level of 90%+ correct prediction of missing values and was ready to be tested in the real world. The first test was carried out in the warehouse, by comparing the predicted sizes of shipments to what was observed on the floor. It was shown that predicted dimensions were over 90% matched by what could be measured with a ruler on the physical shipment. This worked for cartons, irregular-shaped, and palletized shipments. The second test was done at the linehauls (trucks) level. Comparing the loading factors measured by AI-enhanced data versus what could be seen on security camera recordings of trucks departing the warehouse, there was again a match of over 90% observed between AI data and what could be seen on camera.

All in all, the data cleansing process showed how self-evident the implementation of AI in shipping is, but even more so the value of accuracy needed from data in logistics. Once it was demonstrated that the actual loading factors were significantly below the 95% average accuracy rate originally reported by the company, the implementation of AI improved the quality of the data and increased precision in operations.

Sign up for the free insideBIGDATA newsletter.

Join us on Twitter: @InsideBigData1 – https://twitter.com/InsideBigData1

Leave a Comment

*

Resource Links: