Our global freight forwarding network keeps our customers freight moving across the world.

An open-source TMS: the machine room must be open

Today we have an op-ed by Newton Davis, co-founder of Janus, a fintech focused on shipment-level financial clarity for ocean & air freight forwarders, and Raoul Wintjes, head of international road haulage, air freight & digitalisation at DSLV, the Association for Freight Forwarding and Logistics Germany. Somewhere inside every freight forwarder's operation is a system that almost no one outside the industry has heard of - and that almost no one inside it can easily live without. The Transportation Management System, or TMS, is the operational core of global logistics. It is where shipment events are recorded, where carrier connections are maintained, and where the audit trail of a shipment's journey lives. Ask any freight forwarder to abandon theirs, and you will see what can only be described as institutional panic. And yet, despite this centrality, the TMS has largely remained proprietary, locked behind a small number of dominant vendors whose pricing power has grown in direct proportion to the industry's dependence on them. A conversation we recently had crystallized a question that deserves a wider audience: why isn't the TMS open source? What the TMS actually is Strip away the marketing and a TMS is, at its core, a shipment event log. It answers a small number of fundamental questions: What is the shipment? What gates must it pass through? Has it been loaded? Is it on the vessel or aircraft? Has it arrived? Has it reached the customer? It is a sequence of checkboxes attached to a shipment identifier, enriched over time with carrier data, customs milestones, and operational notes. That event log also functions as the connective tissue between a freight forwarder and the carriers, customs systems, and co-loading partners they work with. A TMS that has been integrated with DBH in Germany, or with the equivalent border systems in France, Italy, the US, Hong Kong, and dozens of other jurisdictions, represents years of painstaking technical work. That integration layer is genuinely valuable -- arguably the most valuable thing a large TMS vendor actually provides. But that is also the beginning of the problem. The monopoly problem When a piece of software becomes essential infrastructure (when an industry cannot function without it) its vendor acquires pricing power that operates almost independently of competitive pressure. This is the situation the freight forwarding industry finds itself in today, with a small number of large TMS providers. Price increases of 20%-30% a year have become routine for some vendors, communicated with little justification and absorbed with little recourse. The reaction from freight forwarders has been significant. Privately, many operators express a strong desire to move off their current TMS within the next few years. But the switching costs are enormous: re-training staff, rebuilding integrations, migrating historical data, and requalifying with carriers and customs systems. The very integrations that make a TMS valuable are also the chains that bind its users to it. It is worth noting that platforms, by their nature, tend toward monopoly. Network effects reward scale; the more participants, the more valuable the platform, and the harder it becomes for any individual participant to leave. This is not unique to logistics software; it is a structural feature of platform economics. The question is not whether the TMS market will consolidate, but whether that consolidation must happen under private ownership. Agentic disruption changes the calculus The case for open-sourcing the TMS is made more urgent - and more achievable - by the emergence of agentic AI. At the 2025 TMS Conference in Frankfurt, the thesis that "agentic AI can eat TMS" was articulated openly by industry observers. We believe this characterisation is largely correct, but it requires precise interpretation. The challenge with the modern TMS is not the event log itself. It is everything around it. The data that should flow into and out of that log -- rate confirmations, booking documents, carrier bills of lading, customs entries, demurrage and detention notices, standard operating procedures, exception alerts -- lives predominantly outside the system. It sits in PDFs, in email threads, in WhatsApp and WeChat messages, in Excel files. A human being is responsible for translating all of that information into the TMS. The TMS is not the problem. The interface between the world and the TMS is the problem. This is precisely where AI agents are creating transformative value. An agent that can read a carrier rate sheet, parse a booking confirmation, reconcile a freight invoice against a quoted tariff, and flag discrepancies in real time, and without human intervention, is not replacing the TMS. It is augmenting the interface layer that the TMS has never been able to automate on its own. The value, in other words, is migrating from the event log to the intelligence that feeds it. As this intelligence layer matures, the TMS itself becomes increasingly commodity infrastructure. If the real competitive differentiation in logistics technology lies in the quality of the AI agents and the accuracy of their invoice audits, the speed of their exception management, the depth of their carrier integrations. Then the underlying event log is simply a shared utility. And shared utilities do not need private owners. Events, not contracts: getting the focus right Before proposing a solution, it is important to be precise about what should be open-sourced, because not all logistics data is equal. The contracts between a freight forwarder and their shipper, or between a freight forwarder and their carrier, are proprietary. They represent the commercial relationships, the margin structures, and the competitive intelligence that define a freight forwarding business. No freight forwarder will (or should be asked to) put those on a shared platform. This is partly why the Open Logistics Foundation's focus on the eCMR, while valuable, has not captured the industry's imagination as fully as it might. The CMR contract, while important, is ultimately still a contract: something that companies are accustomed to handling privately. The events are different. Knowing that a container was loaded at Hamburg on the 14th, that it transshipped at Rotterdam on the 18th, and that it was available for collection in Newark on the 28th, is not commercially sensitive in the same way. It is operational data; the kind of data every party in the supply chain (shipper, freight forwarder, bank, insurer, receiver) needs to see, from their own angle, to do their job. This is the data that should be open. The ideal solution combines both layers: a shared event-tracking infrastructure on top of which smart contracts and digital transport documents (including the eCMR) can operate. The event log becomes the authoritative source of truth; the contract layer defines the obligations triggered by those events. Combining them, and making the infrastructure beneath them open, would be genuinely transformative for global trade. The TradeLens lesson The industry has been here before. TradeLens, the blockchain-based shipping data platform launched by IBM and Maersk in 2018, was technically sophisticated and addressed a genuine problem. Freight forwarders that participated in its pilot programmes were, by and large, genuinely excited about what it could do. They just did not trust it. The reason was simple: TradeLens was owned by a carrier and a technology giant, both of which were also competitors or potential competitors of the freight forwarders being asked to share their data on the platform. Sharing operational data with a system controlled by Maersk was, for many freight forwarders, an unacceptable commercial risk, regardless of the technical architecture. TradeLens was discontinued in 2022. The lesson is not that the concept was wrong. It is that ownership matters as much as technology. A shared logistics infrastructure platform can only achieve its potential if no single player can dominate it, extract data from it for competitive advantage, or price others out of participation. A foundation model for shared infrastructure The governance model that makes the most sense for an open-source TMS is the foundation model. A neutral, non-profit foundation would be responsible for maintaining the core platform. Its event schema, its carrier integration libraries, its customs system connectors with a sustainable fee structure to fund ongoing development. Participants would pay usage fees proportional to their activity, not rents proportional to their dependency. Critically, every participant would also be a co-owner. Governance rights, votes on platform development priorities, pricing policies, and integration roadmaps would be distributed across the user base. No single freight forwarder, carrier, or technology company could unilaterally change the rules. This is the core principle that TradeLens lacked and that any successor platform must enshrine. This structure enables a healthy competitive ecosystem at the layer above the infrastructure. If the event log is shared and the carrier integrations are collectively maintained, then the competition between technology providers can happen at the agentic layer,where it genuinely creates value. Who builds the best invoice reconciliation agent? Who has the most accurate exception prediction model? Who can generate the most complete pre-shipment financial picture? These are questions that should be answered by the market. The question of "who owns the event log?" should not be. Why now? The moment for this conversation is particularly well chosen. The freight forwarding industry is in the early stages of a profound renegotiation of its relationship with technology vendors. Pricing pressures from incumbent TMS providers have created genuine willingness to explore alternatives, perhaps the highest level of openness the industry has shown in a decade. At the same time, the maturation of agentic AI is reshaping what "value" means in logistics software, reducing the perceived importance of the TMS core while elevating the intelligence layer around it. The conditions for a serious industry-wide conversation about open infrastructure are better now than they have ever been. The alternative, continuing to pay monopoly rents on essential infrastructure while competitors build proprietary agentic layers on top of that same infrastructure, does not serve the long-term interests of freight forwarders, their customers, or the broader trading system that depends on their efficiency. A call to the industry We are not proposing a finished solution. We are proposing a conversation; one we believe the logistics industry is ready for. The questions are straightforward: what would a foundation-governed, open-source TMS look like? Which organisations would need to be at the table? What is the minimum viable shared standard for shipment event data? How do we structure the carrier integration layer to be collectively maintained? And how do we ensure that the governance model precludes any single party from capturing the platform for their own benefit? These are solvable problems. The technology is available. The incentive is clear. The window is open. What the industry needs now is the will to walk through it.

Strong Q2 for Gemini partners, but Maersk steams ahead over first half

Gemini by name, Gemini by nature, as 'cooperators' Hapag-Lloyd and Maersk presented very different faces with their H1 earnings announcements: the Danish carrier beaming on the back of strong growth; the German line reporting a downturn. Group-wide revenue for the Copenhagen-headquartered shipping giant jumped 8.6% year on year over the six months to June, to $28.7bn, generating profits, measured as EBITDA, of $4.7bn, an upswing of 16.5% compared with H1 25. During an investor call attended by The Loadstar, CEO Vincent Clerc singled out container market demand as the core growth driver, noting it had been "extremely resilient, driven by export growth out of Asia". He added: "This has continued relentlessly despite various events, such as the war in the Middle East or a new round of tariffs, with demand out of Asia up 6.2% in Q2, and our weekly volumes above where they were prior to these events." For the Ocean unit, H1 revenue climbed 6.9% year on year, to $18.7bn, driven by rate spikes in Q2, but profits fell 12%, to $2.9bn. The group's second-quarter figures, praised by those on the investor call, included a 20% revenue surge, to more than $15.7bn, generating profits, of $2.9bn, an upswing of 30% on 2025. CFO Robert Erni said: "We had a good second quarter, with results stronger in comparison with both the prior year and the first quarter, driven by all three [business] segments but in particular Ocean, as higher spot rates and volumes translated into better earnings. "The largest contributor was freight rates, which alone had a positive impact of about $1.6bn on EBITDA. It included compensation for higher bunker costs, elevated insurance premiums, longer dwell times and other costs associated with contingency routing." Such buoyancy was less visible at Maersk's Gemini partner, Hapag-Lloyd, which reported downturns in earnings and profitability over the first six months, blaming a "challenging market environment and operational disruptions". H1 revenue for the German carrier fell 4.7%, to $10.6bn, resulting in a disastrous profit drop of more than 35%, to $1.2bn, thanks to "weather-related disruptions" and the continuing situation in the Strait of Hormuz. But there was a note of optimism from CEO Rolf Habben Jansen, who said: "The second quarter was better than the first, driven by significantly higher spot rates and robust demand. Our Gemini network remained resilient and continued to outperform the market, setting industry benchmarks for schedule reliability. "The terminal business continues to grow and is becoming increasingly strategically relevant, supported by strong throughput and investment in new assets. We will remain focused on growing our liner shipping and terminal businesses while maintaining strict cost discipline." Its Q2 revenues hit $5.7bn, an increase of 8.3%, albeit with profitability holding roughly flat on 12 months ago. at $821m. Nonetheless, the carrier opted to raise its outlook on what proved to be a "better-than-expected" quarter.

Emirates SkyCargo in out of this world transport

Emirates SkyCargo has completed the transport of Altair-1, the UAE's first commercial AI-enabled earth observation satellite, from Dubai to Los Angeles ahead of its scheduled launch in October 2026. The shipment was carried on a Boeing 777 freighter from Dubai World Central (DWC) to Los Angeles (LAX) and was managed under the carrier's Aerospace and Engineering product. The airline said that a specialist team oversaw the end-to-end movement of the satellite, "maintaining strict environmental and safety controls throughout the journey and ensuring delivery within the required launch timeline". Emirates SkyCargo said the shipment was one of a number of satellites and other high-value aerospace equipment that it has transported in recent years. Earlier this year, the carrier transported Arab Satellite 813 from Dubai to its launch site near Shanghai, China. In 2018, it transported Khalifasat, the first satellite developed and built by UAE engineers, from Dubai to South Korea. Emirates said its Aerospace and Engineering product is one of the fastest-growing among its portfolio. In 2026, the carrier has moved more than 6,000 tonnes of aerospace and advanced engineering cargo worldwide. Badr Abbas, divisional senior vice president at Emirates SkyCargo, said the movement of Altair-1 demonstrated the airline's experience in handling complex aerospace shipments. "Transporting highly sensitive and technologically advanced cargo requires precision and expertise, and we are well positioned through decades of deep expertise to support this critical mission," he said. He added that the shipment underlined the role of specialised air cargo solutions in supporting the global aerospace sector. The satellite was transported for UAE-based manufacturer Orbitworks and is the first of 10 satellites planned for the company's Altair constellation.