How AI is improving warehouse performance and easing supply chain disruptions

Unlocking greater performance gains in warehouses using artificial intelligence (AI) and machine learning (ML) helps make supply chains more resilient and capable of bouncing back faster from disruptions. Unfortunately, the severity and frequency of supply chain disruptions are increasing, with McKinsey finding that, on average, companies experience a disruption of one to two months in duration every 3.7 years. 

Over a decade, the financial fallout of supply chain disruptions in the consumer goods sector can equal 30% of a year’s earnings before interest, taxes, depreciation and amortization (EBITDA). However, Fortune 500 companies with resilient supply chains achieved a 7% premium on their stock price and market capitalization. 

Resilient supply chains are the shock absorbers that keep ecommerce, retail, grocery, and post and parcel businesses running despite the quickening pace of disruptions. Hardening supply chains to make them more resilient pays.

Closing warehouse gaps strengthens supply chains

Unexpected delays and undiscovered warehouse mistakes cost the most to fix and wreak havoc across supply chains. Warehouse managers, planners and expeditors rely on decades-old processes based on Microsoft Excel spreadsheets. But, with increasing costs, pace and severity of disruptions, warehouses can’t react fast enough with these manual systems. As a result, “Operations managers are spending hours collecting data and entering it manually into Excel spreadsheets, taking valuable time away from managing and optimizing warehouse operations,” Akash Jain, Honeywell connected enterprise general manager for connected warehouse, told VentureBeat. 

Warehouse accuracy and performance further slow down because decisions made on the warehouse floor that impact margins, costs and revenue trade-offs often don’t make it to the top floor. Senior executives need to know how split-second decisions on which orders to ship impact inventory carrying costs and total inventory value. Runaway inflation makes inventory valuation one of the most expensive risks to manage today.

Stress-testing supply chains often uncovers the largest and most costly gaps in warehouse performance down to the asset level. Asset performance management (APM) must be a core part of managing a warehouse, so the cost, risk and machinery used can be optimized with real-time data. 

For warehouses to absorb disruptions and keep working, the managers running them need a continual stream of near real-time data from supervised ML algorithms to optimize their operations’ many constraints. “Many distribution businesses were caught completely by surprise when ecommerce demand took off at the start of the pandemic. Many were running multiple shifts to keep up with demand, with little to no time to keep machinery and warehouse assets maintained so they wouldn’t break down,” Jain told VentureBeat. 

How AI is closing warehouse gaps 

The more fragile supply chains become, the more important it is to find where warehouse gaps are and close them. By using supervised ML algorithms and convolutional neural networks, it is possible to use the real-time data streams generated from warehouses to pinpoint where gaps are. However, identifying just how wide these gaps are, their impact on daily warehouse operations and their financial impact on a business has proven elusive. 

Cloud-based enterprise performance management (EPM) platforms are taking on that challenge. They’re combining APM with site operations applications to identify how warehouse sites perform against plan, helping managers identify bottlenecks and solve them before they impact performance. Leading EPM providers rely on APIs to integrate with current and legacy warehouse management systems, differentiating themselves by functional area and vertical market. Oracle, SAP, IBM, Anaplan, OneStream Software and Honeywell Connected Warehouse offer EPM platforms today.  

Of the many approaches enterprise software vendors are taking today, Honeywell’s Connected Warehouse platform strategy and use of AI and machine learning are noteworthy. It leads the EPM platform market in using advanced ML techniques and constraint modeling to identify warehouse and logistics bottlenecks. 

AI and ML are designed into the foundation of Honeywell’s Forge platform and portfolio of products. The company has more than 150 AI and data science experts on staff, concentrating on the Honeywell Forge roadmap, future innovations and new patent opportunities. 

All these AI and ML investments translate into continual improvement in providing real-time insights and contextual intelligence that improves warehouse and supply chain performance. The goal is to provide distribution businesses with a real-time system of record they can use to identify gaps in warehouse performance and better manage machinery and assets, said Jain.   

Honeywell’s Connected Warehouse uses ML to analyze real-time data and make recommendations based on constraints while monitoring machinery to see how its performance can be optimized. The dashboard below combines real-time updates for outbound operations, tracking current progress on packed and shipped cartons against the plan. 

Real-time data, analyzed using analytics and ML algorithms, keeps the dashboard current. Constraint-based ML algorithms also calculate planned performance in real time and are used for tracking asset downtime. In addition, Honeywell recently introduced an APM that predicts when warehouse machinery needs preventative maintenance and updates. 

Anticipate more supply chain disruptions 

Stress-testing supply chains needs to start in the warehouse, where small process improvements made at scale can make a difference in keeping distribution centers and networks running efficiently. What’s been missing is a 360-degree view of warehouse performance that can identify how fast bottlenecks are growing and their financial impact. Combining AI, ML, and real-time OT and IT data, cloud-based EPM platforms are taking on this challenge. 

It’s a certainty that more supply chain disruptions are on their way. Using AI and machine learning to optimize warehouse operations will help absorb those shocks. AI- and ML-based warehouse management is a necessity today for high-velocity distribution businesses, including ecommerce, retail, grocery, and post and parcel, to reduce the impact of supply chain disruptions.