BLOG

What’s Wrong With Warehousing Technology Today? Is AI The Solution?

Table of Contents

Subscribe to Our Newsletter

Share on Social

Keith Moore is CEO of AutoScheduler.AI.

getty

Many warehouse management systems (WMS) and warehouse control systems (WCS) vendors have a very concrete, fenced-in approach to the world. These vendors believe that a company’s four walls are their fence, and they lock people out, not wanting to expose them to new technologies. This belief hinders the adoption of technologies that could help the business.

In the same way, many executives within a company keep data siloed within various operations. Valuable supply chain data exists within a WMS, labor management system (LMS), yard management system (YMS) and transportation operations. Large enterprises shipping goods worldwide want openness with data and collaboration across vendors, but the way enterprise systems work impedes data sharing.

Vendors need to share data. Customers must come first. Organizations must report back to their management about what is happening in the supply chain and if any issues will affect customers’ orders.

Where does a data warehouse or data lake fit into this concept of data sharing? A data warehouse is a central repository that stores data from multiple sources within an organization, allowing easy access and analysis to support decision-making. A data lake stores data “as is” without predefining its structure, making it ideal for big data analytics, machine learning and other advanced data processing needs.

Workers within an organization need to generate reports for managers, but siloed information makes it hard to make meaningful reports. There has to be a better way. People can pull the data into the formats they need using a replication database, but this is often dependent on the customer. Vendors should publish the data in a common streaming platform.

Vendors need to ensure that data is available from their systems, and then companies can build a data lake using a common streaming platform to publish the data. However, if data is just put into a data lake, which may be too deep, it can be hard to extract and make sense of it.

When is the right time to upgrade a WMS?

Many organizations use a legacy WMS and often wait until the WMS is no longer supported to upgrade. A good time to upgrade is after looking for opportunities for a return on investment (ROI). Companies can shift to portions of a new version of the WMS without taking a big-bang approach. Companies can also use a warehouse accelerator that optimizes and orchestrates operations. These types of systems dynamically plan what will happen and the best way to run a facility so that you maximize value and reduce costs.

Companies are trying to build AI into their operations to streamline decision-making. The first step in the AI journey is making sure the data is good. Then, companies need to build a data process map to uncover gaps in the data. A WMS coordinates many microflows asynchronously. These microflows can be broken down into the 10 most used operational flows, which will help break down operational silos.

Who do you start with? Start with the operations person who knows technology and the operational elasticity of the building. If I push on this balloon, where does it push out elsewhere? What are the critical constraints driving the operation? Where are the bottlenecks?

Businesses need a data openness strategy. Companies need the data at any time. Then, the data will be mapped to microflows of the current and future state of the warehouse. The goal is to keep the supply chain flowing. Companies need to set up a flexible data structure—this is mission-critical for moving forward.

Where do the different types of AI fit in?

Narrow AI

Narrow AI is the term that IT teams use for machine learning-based AI. This term has been used for the last 10 years across supply chains, business processes and applications. In the supply chain, narrow AI is used in the planning and operations execution.

Generative AI

Generative AI does not help decision-making but provides an easy way to interact with information. For example, GenAI helps with the configuration of space in the warehouse.

The data quality across functional silos in the supply chain is not good enough for Gen AI models as they exist today, except in narrow functions. This means that the value and impact of narrow AI across these systems—which are going to be boutique, custom-built models based on available data—are going to have a much more relevant impact on the business and the dollars and cents and service levels in the next 12 to 18 months.

Predictive And Causal AI

Predictive and causal AI can predict when an operation needs to be performed or something needs to be done, like changing the amount of labor needed for an operation. Causal AI orients itself to a more headless approach with actions going in and operations coming out. For example, a business receives a data stream from robots. Does this stream tell why robots move slower in a particular aisle?

The warehouse needs people, but we must augment them with appropriate technology, whether narrow AI, advanced AI or other technology, to help them do their jobs more effectively.

Many AI tools can make employees’ jobs easier, such as writing a manual using generative AI so that new employees understand exactly what they need to do. Creating manuals saves time and adds value. However, with AI, replacing people within the supply chain is a long way away.

Where is the warehouse world going?

Innovators without a good data strategy will be relegated to the sidelines. Data openness must be an organizational philosophy, and sharing that data more ubiquitously is the key to unlocking the supply chain. Businesses should focus on the modularity of the system around that data. They must move away from monolithic software and instead focus on building the system’s modularity into their operations.


Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?


Follow me on Twitter or LinkedIn. Check out my website

Keith Moore

Keith Moore is CEO of AutoScheduler.AI. Read Keith Moore’s full executive profile here.

Resource Library