In an increasingly competitive global economy, warehouse managers are under pressure to transform their supply chain operations with AI applications. “Today’s warehouse must be smart, and that can only happen by working with the data that sits behind the moving of things,” says Forbes. “Smart warehousing integrates new physical and analytical technologies to realize a host of benefits, including faster problem resolution, improved labor efficiency … [and] the ability to predict and better adapt to business demands,” among others.

Artificial intelligence (AI) must play a key role in modern warehouse management operations. Fortunately, growing warehouse AI applications in the supply chain enable warehouse operators to become more competitive than ever.

Functionally, AI can help solve logistics, workforce, safety, and optimization issues quickly — even as warehouse dynamics change in real time. But while “companies are already using AI in their warehouses and distribution/fulfillment operations … operators view cost, complexity and lack of understanding of how to use AI as key impediments to further investments,” Supply Chain Dive reports.

This article explores what it takes to capitalize on AI applications in the supply chain using an ecosystem of technologies for warehouse use cases. It focuses on four key steps that will put warehouse operators on the right path toward greater operational value through their AI journey.

4 Steps to Capitalize on AI Applications in Supply Chain

1. Identify Potential Warehouse Functions for AI Applications

AI can help warehouse operators mitigate the multidimensional requirements of running a truly modern warehouse. Specifically, AI can augment human decision-making, enabling personnel to focus on their core competencies rather than juggling the complexities of logistics or adjusting to real-time operational challenges in the warehouse.

To begin, warehouse operators must identify the most imminent warehouse functions that can benefit from AI applications in the supply chain. Questions they should ask include:

• What business processes or areas currently consume the most time and resources?
• How can their improvements be measured (which KPIs would track such enhancements)?
• What processes or areas are most likely to be disrupted by changes in warehouse conditions (e.g., seasonality, product mix, increased demand)?
• In which functions do employees currently spend the least time on value-added activities?

After warehouse operators identify the appropriate warehouse functions for AI applications in the supply chain, they can begin exploring the different types of vendors that leverage AI technologies in their solutions and how those technologies can deliver quick time-to-value when applied to their specific warehouse needs.

2. The Applicability of AI Technologies for Warehouse Use Cases

Whereas AI technology spans numerous areas of specialization, many of its elements are readily applicable to warehouse use cases. Analytics often plays a key role in this, and it typically spans several areas:

• Descriptive: What happened?
• Diagnostic: Why did it happen?
• Predictive: Could it happen again, and how frequently?
• Prescriptive: What can be done to either circumvent or promote it?

Although all play key roles in any analytics systems, the former two are where traditional business intelligence systems usually operate with heavy reliance on human input for interpretation. However, the latter two are where AI distinguishes itself.

Many warehouse management systems include substantial rule-based functionality, typically configured by human experts. Once set, they are hardly touched. AI can easily and automatically adjust the parameters of such rules to accommodate changing business conditions without any human intervention. It can easily predict events and prescribe courses of action. For example, the AI in the warehouse can predict demand for a particular item and, in response, easily prescribe adjustments to the allocation of inventory space and the frequency of replenishment to meet the pick demand.

The AI can also prescribe efficient putaway, storage, picking, and replenishment strategies to ensure optimal warehouse operations.

3. Align Warehouse Personnel with AI Applications in Supply Chain

Picking typically makes up the lion’s share of warehouse activities (~70%). Any increase in picking efficiency will translate to more productivity. However, warehouse operators must demonstrate the benefits of AI applications in the supply chain (e.g., in picking) to other stakeholders to secure “buy-in.” That includes warehouse personnel, whose work AI will impact. Indeed, personnel may distrust AI-powered warehouse management systems (WMS) because they are unfamiliar with how the technology works or they are concerned about relinquishing decision-making responsibilities to AI.

Warehouse operators can address these concerns by involving warehouse personnel in the AI adoption process and providing them with training on interacting with AI systems. Furthermore, warehouse operators can communicate the value AI brings to the warehouse, such as improved time management and reduced stress.

For example, AI can maximize workers’ pick productivity during regular shifts to reduce late nights, overtime, and exhaustion. Personnel can also spend less time struggling with logistics complexities and more time focusing on their core responsibilities (e.g., operating a forklift safely and effectively). AI-driven decision-making can help mitigate mental overload during unexpected or rushed events, such as when multiple trucks arrive or depart simultaneously, when personnel would otherwise struggle to identify the best course of action on time.

Initial impressions of AI-based instructions may be negative. For instance, the pathways AI prescribes for a pick path may be counterintuitive to the forklift driver, who is used to S-shaped traversal in parallel aisles and a warehouse setup. But warehouse operators can begin with small-scale AI applications to demonstrate its viability quickly and prove to stakeholders and personnel that AI delivers results. Incorporating feedback from warehouse personnel to optimize AI-driven decision-making will help drive adoption.

4. Identify an Opportunity for Initial Adoption

Begin with the “minimizing pick travel distance” problem, where AI can demonstrate successes quickly before you move on to other applications. Early use cases may include:

• Picking: Automate clustering and real-time route guidance (picking sequence) for warehouse pickers using AI-enabled warehouse management systems.
• Storage: Automate storage using warehouse management systems to track inventory levels and reorder or replenish stock as needed.
• Order Release: Automate the order release process using AI to identify when orders are ready for release, aligning with the transport service schedule while ensuring warehouse efficiency.

As personnel begin working with AI applications in the supply chain, they will build trust in its decisions and in its unique ability to improve its own functionality based on real results. Over time, after seeing positive results themselves, their perception of AI as a “black box” may give way to a true sense of increased control.

About the Author:

Vasco Kollokian is an expert in warehouse management at Tecsys, driving innovation through AI-driven and advanced warehouse execution system initiatives across distribution and analytics, retail, and healthcare.