Weekly highlights

Ocean rates – Freightos Baltic Index

Asia-US West Coast prices (FBX01 Weekly) stayed level.

Asia-US East Coast prices (FBX03 Weekly) decreased 6%.

Asia-N. Europe prices (FBX11 Weekly) increased 1%.

Asia-Mediterranean prices (FBX13 Weekly) increased 3%.

Air rates – Freightos Air Index

China – N. America weekly prices increased 9%.

China – N. Europe weekly prices increased 1%.

N. Europe – N. America weekly prices increased 12%.

Analysis

Even with more container vessels routing back through the Red Sea on Asia – Europe services, and estimations of slumping post-peak demand, carriers have been able to push rates up 10% to N. Europe and 23% to the Mediterranean over the past month.

Prices ticked up 3% last week to $4,375/FEU to the Mediterranean and were about level at $3,595/FEU to N. Europe. Congestion in N. Europe hubs, as well as some lingering backlogs in the Far East may be helping put some degree of upward pressure on rates, with a recent one-day port worker union strike across many ports in Europe as well as other regions, possibly adding to backlogs.

Carriers planned GRIs of $1,000 – $1,500/FEU for these lanes to start the month and additional increases for mid-November. Daily rates so far this week have stayed flat, but the coming days will indicate if carriers go through with the price hikes and if they’ll succeed, with some forwarders skeptical that market conditions will support further increases.

Even with transpacific rates still at peak levels, carriers have GRIs of $2,000 – $3,000/FEU set to start the month for these lanes as well, though here too rates so far this week have not climbed and some forwarders think more increases are unlikely as this trade finally enters slow season. East Coast rates cooled 6% compared to last week’s bump, but are still $200/FEU higher than two weeks ago at about $9,900/FEU, with West Coast prices about level since climbing to the $8,300/FEU range in September.

Asia – Europe rates are climbing relative to September, but from July to September prices fell $2,000 – $3,000/FEU on these lanes, about back to their pre-peak season baseline in mid-May. Meanwhile, the surprising rate resilience on the transpacific – including post-Golden Week, with prices currently about $5,500/FEU higher than before peak season – and the relatively short amount of time left until demand is likely to pick up ahead of Lunar New Year may make it unlikely that transpacific prices will fall back anywhere close to their pre-peak baseline before they climb again.

Another factor that could help prop rates up even as demand cools is the potential port worker union strike at LA/Long Beach. The Local 26 watchmen union, part of the ILWU, has been working under an expired contract since mid-2025 and last week voted to authorize a strike as early as November 1st. Over the weekend though, the union agreed to extend the current contract until 5pm PT on November 4th to allow additional time for negotiations, with federal mediators now involved. A strike by the relatively small union could have an outsized impact on port operations as these workers provide required security services at many of the terminals.

In air cargo, Far East – N. America rates continued to climb last week, increasing 9% to $7.76/kg from the peak season demand bump. This price level – a high for the year – is up 40% compared to this time last year, and is approaching the 2025 peak of $8.00/kg reached early last December. Asia – Europe rates were about level at $4.39/kg last week and are 11% higher year on year, but are well short of highs of about $5.25/kg seen earlier this year, as the market continues to adjust to decreases in e-commerce volumes on this lane.

The post Asia – US air rates already nearing last year’s high on peak season volumes – November 3, 2026 Update appeared first on Freightos.

Resilience becomes expensive when it is treated as a shopping list of buffers: more inventory, more overflow space, more carrier capacity, more alternate lanes, more technology redundancy. A better starting point is diagnostic: how does this logistics system fail, how quickly will we know, and what happens next?

Systems engineering offers another useful discipline: failure mode thinking. Instead of beginning with the question, « How much redundancy should we add? » begin with, « How can this system fail, what happens when it does, and what should the system do next? » That produces a much more precise resilience strategy. Resilience becomes more important as the operation moves toward the Autonomous Logistics Network, because tightly connected and increasingly automated systems can propagate failures faster unless recovery paths are designed in advance.

Make Failure an Input to the Design

Every logistics has failure modes.

A carrier can reject capacity. A port or terminal can close. A distribution center can lose power. A conveyor, sorter, robot fleet, or dock system can fail. Labor can become unavailable. A transportation lane can lose capacity. A software interface can fail. Data can be wrong. A cyber incident can isolate systems. An automated logistics decision can be incorrect.

The objective is not to imagine every possible disaster. It is to identify the failures that matter and understand their consequences.

A failure mode becomes important when it combines likelihood, severity, detectability, and recoverability in a way that threatens business requirements. This is familiar territory for engineering disciplines and increasingly necessary for logistics networks.

Map the Dependency, Not Just the Carrier List

Traditional logistics risk programs often focus on carrier capacity, weather, and geographic exposure. That is useful, but dependencies extend further.

Multiple carriers may depend on the same constrained gateway. Several distribution centers may depend on one cloud or integration service. A parcel strategy may depend on one sort hub. Two apparently redundant transportation options may share the same port, airport, rail ramp, road corridor, or technology provider. A seemingly resilient logistics network can still contain common points of failure.

Resilience therefore requires dependency mapping. What facilities, lanes, carriers, gateways, data feeds, people, automation assets, and technologies are required for a critical logistics flow? Which dependencies are hidden? Which are shared?

This system view helps distinguish real redundancy from apparent redundancy. Two options are not redundant if they fail for the same reason.

Early Detection Creates Options

A failure that can be detected early is different from one discovered after the operating window has closed. That makes visibility part of resilience.

If a carrier-capacity problem is identified days before a critical shipping window, the organization may have choices: re-tender, reroute, reposition inventory, change mode, or use another facility. If it is discovered after the dock cutoff or delivery window has closed, those choices disappear. Failure-mode analysis should therefore ask not only what can fail but how the organization will know.

Which signals provide early warning? How reliable are they? Who monitors them? What threshold creates an exception? What decision follows?

This is where digital threads, control towers, and AI can make resilience operational rather than retrospective. The value is not seeing disruption. It is seeing it while options still exist.

Design the Degraded Mode Before the Crisis

Resilient systems do not always continue operating normally during failure. Sometimes they operate differently.

A warehouse automation failure may require a reduced manual mode. A TMS outage may trigger frozen routes and predefined carrier instructions. A carrier-capacity loss may activate backup providers or alternate modes. A transportation disruption may activate alternate lanes or facilities with higher cost and lower capacity. These degraded modes should be designed before the event.

These degraded modes should be designed before the event.

What service can still be maintained? Which customers or products receive priority? What manual processes are available? How long can the system operate that way? What resources are required?

The alternative is improvisation during crisis. Some improvisation will always be necessary. It should not be the entire resilience strategy.

Buy Resilience Where the Failure Matters

Resilience has a cost. The organization may pay for overflow warehouse capacity, backup carriers, alternate lanes or ports, premium transportation options, spare automation components, backup connectivity, cybersecurity controls, or additional monitoring. The engineering challenge is to match the investment to the risk.

Not every facility, lane, carrier relationship, or technology component needs the same protection. Not every customer flow needs the same recovery time. Criticality should determine the logistics resilience requirement.

This is where scenario analysis and digital twins can help. Organizations can test the consequences of failure and evaluate which combination of mitigation measures produces the best risk-adjusted result. That is more disciplined than simply adding buffers everywhere.

Near Misses Are Free Failure-Mode Data

Logistics networks experience countless small failures that never become major disruptions because somebody intervenes. Those near misses are valuable data.

Why did the carrier issue not become a service failure? Why did the warehouse recover from the system outage? Why did the customer order remain on time despite a transportation delay? Was the recovery designed or accidental?

Organizations should capture these events and feed them back into system design. A near miss may reveal a hidden dependency, an effective workaround, a weak interface, or an employee carrying risk through heroic effort. If resilience depends repeatedly on heroics, the system is telling you something.

Engineer Recovery, Not Just Redundancy

The strongest resilience strategy is not the one with the most backup capacity. It is the one that understands how the system can fail, detects important failures early, contains their consequences, and recovers deliberately.

The strongest resilience strategy is not the one with the most backup capacity. It is the one that understands critical dependencies, detects failure while options still exist, and knows how to degrade and recover deliberately. The final article pulls the full series together into the idea of continuous systems engineering.

Related Logistics Viewpoints research

Systems Engineering in Logistics
The New Architecture of Logistics
2026 Autonomous Exception Management Market Map
Three Ways to Design Your Supply Chain for Successful Mergers and Acquisitions
Previous in this series: Govern the Living System: Change Control and Lifecycle Management

The post Design for Failure: Resilience, Risk, and Failure Modes appeared first on Logistics Viewpoints.

Artificial intelligence has become the defining technology conversation of the decade. Every organization is exploring how AI can improve productivity, automate processes, and unlock new business value. Yet for supply chain leaders, the opportunity extends far beyond automation. AI has the potential to fundamentally change how organizations anticipate disruption, make decisions, and orchestrate outcomes across increasingly complex global supply networks.

However, there is a critical reality that many organizations are discovering: AI is only as powerful as the data behind it.

Without access to accurate, connected, and governed data, even the most advanced AI models produce incomplete insights and unreliable recommendations. For AI to deliver measurable business value, organizations must first solve one of the biggest challenges in supply chain management: bringing together fragmented data into a trusted, unified foundation.

This is where InterSystems delivers a unique advantage.

The Supply Chain Has a Data Problem

Modern supply chains generate enormous volumes of data every second. ERP systems manage orders and financials. Warehouse management systems track inventory. Transportation systems monitor shipments. Manufacturing execution systems oversee production. Supplier portals, IoT devices, clinical systems, weather feeds, and third-party logistics providers all contribute valuable information.

The problem isn’t a lack of data. It’s that the data lives in hundreds of disconnected systems. When disruptions occur, organizations often spend hours (or even days) collecting information before they can make a decision. Teams manually reconcile spreadsheets, validate conflicting reports, and coordinate across departments to understand what is happening.

By the time they reach a conclusion, the situation has often changed. AI cannot solve this problem if it only has access to fragments of the story. It requires a complete, trusted view of the enterprise.

Unified Data Is the Foundation for AI

InterSystems believes AI should not operate in isolation. It should be embedded within operational workflows and powered by real-time, trusted enterprise data.

Rather than requiring organizations to replace existing systems, InterSystems connects information across the entire supply chain, creating a unified, governed orchestration layer that makes data accessible wherever it resides.

This enables organizations to:

Eliminate data silos without large-scale migrations.

Deliver trusted, consistent data to AI models.

Provide real-time visibility across the supply chain.

Govern and secure enterprise information.

Create a single source of truth for operational decision-making.

With a unified data foundation, AI shifts from being an experimental technology to becoming a trusted decision-making partner.

Moving Beyond Visibility to Decision Intelligence

Visibility has long been a priority for supply chain leaders. Dashboards, reports, and analytics provide valuable insight into operations, but visibility alone doesn’t improve outcomes.

Organizations don’t need more dashboards. They need faster, better decisions. And they need to be able to understand not only what is happening, but what is likely to happen next and what actions will produce the best business outcome.

Instead of simply identifying that inventory is running low, AI can recommend how to rebalance inventory across distribution centers. Instead of highlighting a supplier delay, AI can identify alternative sourcing options before production is affected. Instead of reporting missed service levels, AI can recommend corrective actions before customers experience disruptions.

This evolution from descriptive analytics to decision intelligence is where AI delivers its greatest value.

The Future Is AI-Enabled Decision Intelligence

As supply chains become increasingly interconnected and dynamic, organizations need more than automation. They need intelligence that continuously analyzes operations, identifies emerging risks, and orchestrates the best possible outcomes.

This future depends on more than large language models or predictive algorithms. It depends on trusted data, interoperability, governance, and seamless integration across the enterprise.

These are the capabilities InterSystems has delivered for decades.

By combining a unified data platform with AI-enabled decision intelligence and supply chain orchestration, InterSystems helps organizations transform fragmented information into actionable insight, enabling faster decisions, greater resilience, and measurable business value.

Introducing InterSystems AI Assistants

InterSystems has launched AI assistants for its two supply chain products: Supply Chain Orchestrator and Data Studio with supply chain module.

InterSystems Supply Chain Orchestrator® AI Assistant provides natural language support and allows business users to interact with features supported in Supply Chain Orchestrator. It provides help with three general capabilities.

Accessing supply chain data. This can be a direct query of a specific data entry like a customer order, or a query on aggregated data such as revenue by region and timeline.

Decision assistant. This helps users understand the business logic/rules used for actionable insights through Supply Chain Orchestrator issue analysis.

Memory management. This helps the user understand a specific industry better or minimize user prompts by keeping user-specific preferences or settings, for example. This capability includes short-term and long-term memory management and global and user level memory management.

Supply Chain Orchestrator provides specialized AI agents that augment human decision making and support intelligent business processes. Data agents allow both technical and business users to discover, access, prepare, enrich, and contextualize information from across the enterprise.

Business process agents monitor operations, identify exceptions, coordinate workflows, and recommend or execute actions according to customer-specific business policies, rules, and operational processes. And memory and context agents maintain business context, organizational knowledge, historical decisions, customer-specific operating practices, and industry expertise.

InterSystems Data Studio™ AI Assistant is a low-code generative AI extension for Data Studio that adds intelligent assistants, out-of-the-box agents, and a multi-agent framework. It is designed to help organizations easily explore, analyze, and query structured and unstructured data and visualize answers using a natural language conversational interface.

AI Assistant is designed to work across all of the data that organizations already have. By leveraging the underlying capabilities of Data Studio to connect, harmonize, govern, and manage information across operational applications, databases, data warehouses, cloud services, and other data sources, organizations can create and maintain a unified, consistent, and governed data layer that AI Assistant can leverage to generate more meaningful insights. By giving business users, analysts, applications, and AI assistants access to a single trusted and governed information foundation, organizations can reduce complexity and increase confidence in the accuracy of AI-driven insights and answers.

AI Assistant provides both out-of-the-box agents and a multi-agent framework that enables organizations to build custom assistants capable of orchestrating multiple agents to address domain specific use cases, with custom data sets, rules, and permissions. Rather than relying solely on a single request-and-response pattern, multiple agents can work together to perform more sophisticated analysis and reasoning. For business users, this creates a more accurate and intuitive way to get answers and explore information across business documents and applications. For organizations, it provides an AI foundation capable of supporting more complex queries, delivering more accurate insights, and enabling deeper analysis across enterprise information.

Final Thought

The companies that succeed in the AI era will not simply have access to more data. They will be the organizations that can trust their data, connect their operations, and turn intelligence into action.

That’s where AI delivers its greatest value, and where InterSystems helps organizations lead.

Chris Cunnane is the Global Product Marketing Manager for Supply Chain at InterSystems. In this role, he is responsible for developing and executing marketing strategy and content for the InterSystems supply chain technology suite. Chris has 20+ years of supply chain expertise, leading the supply chain practice at ARC Advisory Group, as well as holding various sales, marketing, and operations roles in the wholesale, retail, and automotive parts markets. He holds a BA in Communications from Stonehill College and an MA in Global Marketing Communications from Emerson College.

The post AI in Supply Chains: Why Trusted Data Is the Competitive Advantage appeared first on Logistics Viewpoints.