LogiPharma Europe: A New Model for Cold Chain Decision-Making

Pharmaceutical cold chain logistics is entering a phase where data quality and decision speed matter as much as physical infrastructure. As temperature-sensitive therapies such as biologics and cell and gene therapies expand, the industry is increasingly relying on sensor-driven visibility and analytics to match packaging systems to highly variable transport conditions.

In her interview with Pharmaceutical Commerce following her presentation at LogiPharma Europe 2026, Raquel Vazquez, global head of Distribution Technology at Roche, outlined how cold chain decision-making is shifting from experience-based selection to data-driven optimization. Vazquez emphasizes that the core challenge is no longer access to data, but the ability to contextualize it across fragmented logistics networks. This includes integrating lane-specific risks, environmental conditions, and product stability profiles into a unified framework that can guide packaging selection in real time.

The complexity of this challenge is amplified by the structure of global pharmaceutical distribution networks, where multiple logistics partners, sensor types, and data standards must be reconciled. Yet Vazquez points out that without contextual connectivity, even high-quality sensor data cannot reliably inform decisions about when to deploy active systems versus lighter, more sustainable packaging alternatives.

Looking ahead, the conversation turns to the evolution of “decision intelligence” in cold chain operations. Vazquez describes a future in which AI and digital twins move the industry from reactive monitoring to predictive intervention, enabling automated recommendations for packaging selection based on real-time and historical performance data. However, she also underscores that human oversight will remain essential, particularly in defining the strategic frameworks that govern these increasingly automated systems.

PC: When using data to pick a shipping container, which specific factors—like outside air temperature or the length of the trip—are the most important for deciding which packaging is the right fit for a certain route?

Vazquez: The goal is to match a shipping system's performance to the specific "thermal challenge" of a route. Over-engineering leads to unnecessary costs and CO2 emissions, while under-protection risks product loss.

Key factors for selection include:

• Understanding your Lane: Granular Lane Risks, from the total transport time, identify the "unprotected time" where a product is exposed to ambient conditions, such as tarmac time, ocean reefer power-offs, or lack of cold storage during customs.
• Understanding the Ambient Temperature profile: Analyze temperature fluctuations across seasons and latitudes; a summer in Abu Dhabi presents a very different challenge than one in Reykjavik.
• Understanding your product: A product with high stability might allow for less protective packaging than one that is highly sensitive to temperature.

This data-driven approach enables fit-for-purpose optimization, allowing us to shift from heavy active containers to lighter solutions like thermal blankets when external temperature conditions permit.

Since shipping involves many different partners and types of sensors, what is the biggest challenge in bringing all that information together so it can be used to make better decisions about packaging?

The biggest challenge isn't just technology or data—both are already available. The real hurdle is creating a framework that provides the connection and context required for true decision making. The three primary obstacles:

• Data Fragmentation and Data Quality: There is a major technical challenge in automatically collecting clean, real-time temperature and geolocation data across a fragmented supply chain into a single, unified data lake.
• Contextual Connectivity: While data exists, it lacks the framework and context. This framework is needed to connect historical shipping performance, weather forecasts, and lane risks to be able to select the right thermal shipping system.
• Lane Risks: Detailed lane visibility is difficult due to diverse partners (LSPs) and constant routing changes e.g. conflicts in specific countries.

From a data perspective, what are the early warning signs that a container might fail to keep its temperature before it reaches its destination?

Today, smart sensors with real-time connectivity allow for active temperature monitoring and intervention before a failure occurs, by detecting when internal temperatures rise. These systems push proactive alarms to logistics service providers when specific temperature thresholds are breached, which may indicate that a parcel was left in uncontrolled storage by error, for example.

As these data tools get smarter, do you think we are heading toward a future where a computer program, rather than a person, will automatically choose the shipping container based on the latest data for that route?

We are definitely moving toward a future of decision intelligence, where systems automatically suggest the most "right-sized" container based on historical performance, real-time forecasts, and Digital Twin simulations. However, this future still requires a human-in-the-loop to provide the strategic framework and oversee the process.

The goal is to move from "what happened" to "what will happen," using data and automation to empower humans to make more proactive, informed decisions.

Different shipping materials react differently to temperature spikes. What metrics do you look at to judge their effectiveness, rather than just looking at whether the product stayed within its temperature range?

To evaluate the effectiveness and performance of thermal shipping systems, we follow the following validation approach

• Operational Qualification (OQ): We conduct physical testing in thermal chambers to prove a system can maintain the required temperature under worst-case, industry-standard profiles, such as ISTA 7D.
• Performance Qualification (PQ): We perform real-life testing to ensure the system performs as expected under actual transit conditions.

With technology rapidly improving, where do you see the cold chain data sector evolving in the next 5 years? Will AI play a role in data analytics and thermal system selection?

In the next five years, AI and Digital Twins will play a key role in cold chain transportation, the cold chain will shift from reactive monitoring (what happened) to predictive action (what will happen).

• Dynamic Selection: By analyzing real-time weather, shipping performance, and product stability budgets, it will be possible to select the most cost-effective and sustainable solution (passive, active, or hybrid) for a specific lane.
• Virtual Qualification: We will move from physical-only testing to Virtual Validation. Digital Twins will simulate "worst-case" scenarios, allowing us to qualify new lanes and systems in hours rather than weeks. Physical testing (OQ/PQ) will evolve into a final verification step rather than the primary method.
• Active AI Intervention: Instead of simple alerts, "agentic" systems will calculate remaining stability budgets in real-time and coordinate with LSPs to take corrective action before a failure occurs.

It is an exciting time to be in this sector; as our data tools get smarter, the potential for massive CO2 reductions becomes a reality, allowing Roche to bring life-changing medicine safely to all patients at the lowest cost for the planet.