The Background to Machine Learning and Artificial Intelligence in Logistics Management

Machine Learning (ML) and Artificial Intelligence (AI) have been used by some larger cargo and logistics companies for forecasting and demand planning for a number of years.  More recently, however, there is increasing interest from businesses of all sizes, particularly as more affordable modular options become available and more Software as a Service (SaaS) suppliers enter the market, or bigger IT players, such as IBM, Google to name a couple, expand their offering.

The technology is developing so quickly, it’s easy to get confused about what is possible with algorithms that learn from historical data produced within companies and integrate wider world experience and apply this to current scenarios to feedback information for data managers to improve real-time decision making.

Definitions change over time, making understanding of what is involved more difficult. DemandPlanning.com offer a definition of AI as: “ …machines being able to carry out tasks in a way that we would consider “smart”…  Now it deciphers unstructured text analytics in almost real time, tells us directions, makes predictions on our purchase preferences, and even talks to us.”

In the past, the expanse of data needed for demand planning has been too complex for machines to process. Nowadays, more advanced, automatic mathematical calculations involving big data happen at ever faster speeds and can do so repeatedly for sensitivity testing, where individual variables are altered in a scenario to anticipate impacts.

Currently we are seeing the point at which machines are capable of surpassing their human counterparts in terms of collating and analysing big data and are able to collaborate with decision makers by being able to highlight potentially significant data, based upon the user’s criteria in order to anticipate trends in business operations, as much as comparing outcomes of applying different tactics, allowing management to make more sound decisions faster. Where complex cargo and logistics management considerations are concerned, this can mean significant strategic and tactical changes at a moment’s notice in response to rapidly changing conditions on the ground.

The Institute of Business Forecasting (IBF) conducted research in 2017 to find out from demand planning and forecasting professionals how they saw their future role in 2025, in view of technological advancements and their impacts on people and process, and technology.  Opportunities for redeployment of staff to more productive and pro-active roles was one of the more interesting impacts for staff of innovative companies embracing the internet of things (IoT). Speed, accuracy, efficiency, cost savings, flexibility are some of the most important benefits highlighted by commentators on how tech is transforming business processes.

As the IoT streams live data from manufacturers’ machines, aggregates online purchase statistics in real time and mobile and communicates with remote devices, transmitting data on demand and making supply chain management and transport logistics more flexible and agile.  Increasingly, computer capacity is empowering freight forwarding companies, brokerages and shippers to respond more effectively to both crisis and business opportunity, ultimately impacting customer satisfaction and much more.

Analytic algorithms in custom made logistics solutions generate data according to the priorities of the user.  This allows managers to impose constraints on resource access remotely, as well as to determine which information is worthy of crunching, for instance, in a cloud-based “demand signal repository, or other data repositories for later use”, says Charles W. Chase, Jr. in the Journal of Business Forecasting (Winter 2017-2018, ‘Real-Time Demand Execution. Anticipating Demand at the Edge’. Vol. 36, Issue 4).

This immediacy of real-time, automated data access, strategic filtering and analysis is called ‘edge analytics’ within machine learning, which is hugely speeding up decision-making processes, as relevant data based on a user’s priorities is collected, filtered and illustrated in flexible report formats for managers and stakeholders to see patterns and anomalies at a glance.

According to Chase, this makes the difference in terms of “…conducting real-time demand execution anticipating demand at the edge…”  For logistics management and supply chain management, the possibilities are seemingly endless, from improved energy consumption, through flexible capacity management, to maximising transport mode and staff resource to support more R&D.

Smart software connected to asset tracking devices is gearing transport and logistics for the future.  From small to medium sized T&L enterprises, through to multi-nationals, where complexity of supply chain management and evolving regulatory environments is a daily reality, custom made logistics solutions are the best investment a company can make outside of fleet and personnel if your technology partner has expertise in this increasingly complex industry.