Automation & Control / Supply Chain

A dose of AI for pharma logistics optimization

What to consider when you’re considering ocean shipping

By Arnaud Morvan, Senior Engagement Director, Aera Technology

For many years, air freight has been the dominant transportation mode for pharmaceutical manufacturers to move products internationally. But now, some pharma companies are actively choosing slower but far less expensive ocean freight over air shipping.

Cost savings are a prime reason behind the shift from air to sea shipping. Millions of dollars in shipping costs are on the line in a pharmaceutical logistics market sized at $82 billion, $15 billion of which is committed to cold chain logistics.

It is estimated that pharmaceutical sea shipping will rise from about 20 percent today to 75 percent within the next 10 years. AstraZeneca is among the leaders in embracing sea shipping, with a roadmap to increase its ocean freight volume from just 5 percent of shipments to 70 percent in the future.

Multiple factors need to be considered on a per-product basis to decide whether ocean shipping is viable, including:

The need for speed. Many urgent-care and limited shelf-life products have a clear-cut need for faster air transport, yet others do not. Instead of air shipping as the de facto transport mode, pharma companies are sizing up implications of sea freight for more commoditized products.

Existing inventory vs. demand. Whether to ship by air or in many cases depends on simple supply and demand: What’s the inventory and safety stock level at a given warehouse, vs. regional demand? But it can be difficult to determine those metrics in real time across a vast global supply chain, especially given long product lead times.

Risks to products. Many cold chain products need to be stored at certain temperatures during shipment, and by regulation can't go to market if they experience temperature deviations beyond specified thresholds of degree and duration. It's a matter of assessing cold chain metrics by sea vs. air and making risk-conscious decisions.

Mark Edwards, a pre-eminent expert on pharmaceutical logistics and AstraZeneca executive, calls the ongoing air vs. sea debate “a boxing match" that will never have one clear winner. Instead, it’s a matter of making the right choices for the right products at the right time.

Applying AI for data-driven decisions


Artificial intelligence is emerging as a powerful way to make the right decisions across the multiple dynamics and data sources that drive pharma logistics — and the transportation industry as a whole.

Consider: Nearly two-thirds (65 percent) of transportation-focused executives believe that logistics, transportation and supply chain processes are in the midst of a “profound transformation,” according to joint study by Forbes Insights and Penske. The report notes, “But of the most visible forces of change, perhaps none carries more potential for innovation and disruption than the evolution of artificial intelligence (AI), machine learning (ML) and related technologies."

AI and ML power cognitive automation, which has the capacity to deliver game-changing improvements in transportation and logistics with data access and compute power far beyond what can be accomplished through traditional tools and human decision-making.

Logistics for Life Sciences, an industry thought leader, foresees significant value form AI in the pharmaceutical transportation space: “AI and predictive software have huge implications for the shipping industry," the publication wrote. “Datasets from pharmaceutical distribution make the application of AI ideal for the industry.”

Unlocking value from transportation data

Beyond deciding on air vs. sea shipping, pharma also faces obstacles in domestic transportation. For example, what’s the fastest, most cost-effective means of getting product unloaded at an airport or sea port to thousands of hospitals and pharmacies in a given region?

As it is, transport managers have limited visibility into the many supply chain dynamics that affect logistics performance. On-hand inventory, demand spikes, carrier availability, capacity, locations and more go into the logistics equation. Those data points are typically scattered across multiple internal and external data sources.

Crucial data isn’t instantly accessible when a shipping disruption occurs, while transporters contend with volatile freight rates, fuel costs, port backlogs, highway construction and other variables that make up a large chunk of any company’s cost of goods sold (COGS) metric.

Cognitive automation supports informed logistics decisions in areas including:

Transport capacity. AI makes it possible to predict and manage transportation capacity at a highly granular level, while virtually eliminating manual work and best-guess decisions. AI accounts for the full set of constraints, such as availability of trucks, containers and drivers in a given area, volume to deliver and available-to-promise (ATP) delivery schedules.

Transport lead times. AI monitors and adapts lead times needed due to volumes, wait times by ports, distribution centers, national borders, highway congestion or equipment failure. And it can account for destination type — for instance, delivery to a single-family home is typically faster than to a business location — or a shortage of drivers or vehicles in a given area.

Real-time decisions. AI insights and recommendations present clear options make the best decisions in difficult situations. For instance, if a shipment is stalled at a warehouse, do you expedite transportation (extra cost), or miss service level metrics (revenue loss, potential penalties)? Granular details and AI recommendations help managers choose the lesser of two evils.

End-to-end supply chain. AI-supported transportation shouldn’t exist as a silo. Cognitive automation ties transportation and logistics to supply chain processes such as demand forecasting, production and inventory management. That gives logistics teams early warning of upstream disruptions that could impact downstream delivery schedules, and it supports more proactive supply chain management.

Amazon, which recently bought the online pharmacy PillPack, is a great example of a company that’s embedding AI in its end-to-end supply chain, “from the website to the warehouses to the actual delivery to your doorstep,” as the NPR report put it.

Now with the Amazon PillPack service, pharma logistics may feel the “Amazon effect” rising consumer expectations and industry disruption. As a countermeasure, pharma companies that embrace AI for logistics can expect significant payback in speed, cost-efficiency and risk mitigation. Those that stick with the status quo may find themselves buffeted on stormy seas or dispatching delivery trucks down a dead-end street.