TradeLens was a blockchain-based supply chain management platform co-built by Maersk and IBM. Launched in 2018 on Hyperledger Fabric, shut down in December 2022 after failing to achieve the industry-wide adoption needed for commercial viability.
- Technology stack: Hyperledger Fabric (permissioned blockchain), IoT sensors, smart contracts for automated document processing
- Pilot results: 70–90% reduction in documentation costs; US freight transit times cut by 40%; container status query steps reduced from 5–10 to 1–2
- Peak scale: 175+ organizations, 600+ ports and terminals, over 50 million tracked supply chain events
- Why it shut down: competing carriers refused to share operational data on a platform co-owned by their direct competitor Maersk; the governance model failed to establish neutral industry trust
- Key lesson: enterprise blockchain in industries with multiple competitors requires a neutral governance structure — no incumbent competitor can own the shared infrastructure
What TradeLens Was — and Why This Case Study Still Matters
In early August 2018, IBM and global shipping company A.P. Moller-Maersk Group officially launched a cargo tracking and supply chain management platform called TradeLens (previously known as Global Trade Digitization, or GTD). At the time of launch, it was the most ambitious enterprise blockchain project in logistics history — and one of the most instructive in its failure.
This case remains an essential reference for any CTO or product director considering blockchain as an infrastructure solution for multi-party business processes. Not because TradeLens succeeded — but because it ran the full cycle: from pilot to scale, from scale to shutdown — leaving behind documented metrics, architectural decisions, and governance mistakes that are directly applicable today.
In this article: the technical architecture of the platform, pilot results, the reasons behind the 2022 shutdown, and practical lessons for anyone building distributed ledger solutions today.
Context: Why Shipping Needed Blockchain
Maersk. A Danish transnational conglomerate specializing in container shipping. Since 1996, the world's largest operator of container ships: over 88,000 employees, offices in 130 countries, and more than 15% of global TEU market share. After a 2016 restructuring, the transport and logistics division took on over 600 vessels with a combined gross tonnage exceeding 3.8 million registered tons.
That scale is equivalent to the supply chain of an entire city. And it was precisely that scale that made effective management through existing systems impossible:
- Cumbersome information and financial transactions between freight and supply chain participants
- An excess of manual, labor-intensive processes generating errors and cargo delays
- Opaque customs processes creating conditions for corruption and fraud
- Unequal information sharing between supply chain participants
- Lack of industry standardization leading to inconsistent and low-quality data
One telling figure: a single cargo shipment generated up to 200 communications between 30 different organizations — carriers, banks, customs agencies, insurers. The paper trail from one flower shipment between Mombasa and Rotterdam stood 25 centimeters high.
IBM Corp. The American IT giant with $9 billion in net income and 350,000 employees across 177 countries. IBM's blockchain division drove the company's distributed ledger efforts. In March 2017, Maersk and IBM announced a partnership and joint venture to commercialize a solution built on permissioned Hyperledger Fabric blockchain — one of the Hyperledger projects developed under the Linux Foundation.
TradeLens Technical Architecture
TradeLens was positioned as an open, neutral platform for managing ocean freight and supply chains on a blockchain foundation. Its core premise: automated information exchange between all supply chain participants without intermediaries or verifying third parties.
Three Ecosystem Layers
Business Network. Shippers and consignees, freight forwarders, ocean carriers, ports and terminals, customs brokers, government agencies. Each organization inputs data into the ecosystem where it is stored, processed, and tracked throughout the cargo journey.
Participation levels within the ecosystem:
- Fully integrated participants — connect via direct API and interact with the platform in real time
- EDI channel participants — use existing EDI standards to exchange data with TradeLens
- Portal participants — interact via web interface without direct system integration
Blockchain Layer. Implemented on Hyperledger Fabric — a permissioned blockchain where every network node belongs to a verified participant. The blockchain acts as the "source of truth": it guarantees that information entered into the system is authentic and error-free. In other words, TradeLens itself is the data validator.
Marketplace and API Environment. An open API environment allowed third parties to develop and launch their own applications on top of TradeLens. A sandbox provided a way to test integrations without risk to production data.
How It Works: IoT + Smart Contracts
The technical logic of TradeLens was built across three interaction layers:
- IoT layer. Sensors attached to cargo, transport, and port equipment track cargo location and condition (temperature, humidity, vibration) in real time. Sensor data is written to the blockchain and becomes available to all relevant parties.
- Smart contract layer. In pre-defined scenarios, sensors communicate with each other through smart contracts. For example: when a container's sensor comes on board a vessel, it automatically transmits cargo information to the ship's system. Upon arrival at port, the vessel automatically transmits full cargo details to port systems.
- Document automation layer. IoT data and smart contracts drive automatic document completion: financial calculations, customs duty payments, audit records. This automation was the primary mechanism for reducing workflow time and transaction costs.
The choice of permissioned blockchain was deliberate. In shipping, each participant needs role-based data access: the shipper sees their cargo, customs sees the shipping documents, the bank sees only the financial transactions. A public blockchain (Ethereum, Bitcoin) cannot provide that level of access granularity and does not scale to global shipping volumes.
Hyperledger Fabric addresses this through channels — isolated sub-networks for bilateral or multilateral interactions, each with its own ledger, smart contracts (chaincode), and permissions. This allows, for example, a bank and a carrier to exchange letters of credit without a competing carrier having access to that data.
But this same architecture contained the seed of the eventual failure: the ordering service in Hyperledger Fabric is a centralized component. At sufficient scale, it becomes a single point of failure and requires complex governance across independent organizations, each with their own security policies — something TradeLens never solved.
The Pilot Project: Flower Delivery from Mombasa
Until August 2018, TradeLens spent more than a year in beta testing. The pilot project chosen was a flower delivery chain from Africa to Europe.
First, Maersk analyzed a single flower shipment sent from Mombasa to Rotterdam in 2014. The result: that one delivery generated over 200 communications between 30 different organizations — producers, logistics companies, banks, and government agencies. The physical paper trail from that single shipment was 25 centimeters tall.
Maersk and IBM then connected all participants in that delivery to TradeLens and sent a new batch of flowers. As soon as the producer submitted a packing list via PC or mobile device, that action was recorded on the blockchain — and simultaneously a smart contract was initiated. The smart contract accompanied the cargo throughout the entire supply chain, automatically filling out documents, applying stamps, and executing financial calculations.
Measured Results After One Year of Platform Operation
| Metric | Before TradeLens | After TradeLens | Change |
|---|---|---|---|
| Documentation costs (avocado, Mombasa → Rotterdam) | $300 (15–20% of delivery cost) | $30–90 | −70–90% |
| US freight transit time | Baseline | 40% reduction | −40% |
| Steps to answer "where is my container" | 5–10 steps | 1–2 steps | −80% |
| Communications per shipment (pilot) | 200+ across 30 organizations | Automated via blockchain | Manual interactions eliminated |
From Our Practice: Blockchain in B2B Trade Is Always Harder on Mainnet
From our experience building B2B financial platforms with multi-party transaction tracking: the most common technical mistake in enterprise blockchain projects is assuming that a pilot on an isolated environment equals production.
In real deployments, different blockchain network nodes have fundamentally different sync times and behavior under load. If the system confirms a transaction in seconds on a test environment, production behavior with real network traffic and competing transactions can look completely different. For logistics systems where a cargo checkpoint may be on a vessel in the middle of the ocean with an unstable connection, this is an architectural risk that needs to be designed in from day one — not discovered after launch.
A separate challenge is off-chain data integrity. Sensor data (temperature, GPS coordinates) does not physically live on the blockchain — only its hash is stored there. If an IoT device or oracle sends incorrect data, the blockchain will faithfully record it as "verified". TradeLens addressed this through a network of trusted data providers — which was, in practice, re-introducing centralization into a system whose value proposition was decentralization.
Partners and Participants at Launch
At the official August 2018 launch, the TradeLens ecosystem included 92 organizations. Key participants:
- Ports and terminals: Modern Terminals (Hong Kong), Patrick Terminals, PSA Singapore, ports of Halifax, Rotterdam, Bilbao, PortConnect (New Zealand), PortBase (Netherlands) — over 20 operators in total
- Ocean carriers: Pacific International Lines (PIL), Hamburg Süd, and Maersk Line as global container carriers
- Logistics companies: DAMCO, WorldWide Alliance, Agility, Kotahi, CEVA Logistics, and others
- Customs authorities: Singapore, Saudi Arabia, the Netherlands, Peru, and Australia
Later additions included Mediterranean Shipping Co (MSC), CMA-CGM, Zim Integrated Shipping Services, Procter & Gamble, and Russia's Ministry of Transport. At peak, the platform connected over 175 organizations and 600+ sea gateways worldwide.
Notably, CMA-CGM joined TradeLens without disconnecting from the competing GSBN platform — a clear hedge, not a commitment. TradeLens's market share at launch: ~4.86 million TEU (21% of the market) versus GSBN's ~7.4 million TEU (32%). CMA-CGM's dual participation pushed TradeLens to a 32.5% share — but the underlying adoption problem remained unresolved.
Why TradeLens Shut Down in 2022
In December 2022, Maersk and IBM formally announced the closure of TradeLens. The official language: "failed to achieve the necessary commercial scale". The real causes were structural — and they are instructive for any enterprise blockchain initiative.
Cause 1: Governance Failure — A Competitor Cannot Own Neutral Infrastructure
This is the central and most important mistake TradeLens made. A platform designed to unite the entire container shipping market was controlled by one of the largest players in that same market. MSC, CMA-CGM, and other major carriers were entirely rational in their hesitation: sharing real operational data on a Maersk-controlled platform meant voluntarily handing competitive intelligence to a direct rival.
No technology fixes this problem. Even if Hyperledger Fabric had guaranteed perfect data isolation across all scenarios (which it does not) — the reputational and competitive risk of associating with a Maersk-owned platform was real and persistent.
From our experience building multi-party B2B platforms for international settlements — where every transaction involves a sending party, a receiving party, and an agent — neutral ownership is not an optional feature; it is a baseline architectural requirement. No participant should have privileged access to another participant's data.
For TradeLens, the correct path was one of two options: (1) an independent consortium with equal representation from all major carriers and no controlling shareholder with competitive interests, or (2) a fully decentralized architecture with no single owner — where the code is the law, not corporate agreements. There is no workable middle ground.
SLA timers, escalation logic, and anti-tamper document verification must all be equally transparent and accessible to all participants simultaneously. When one participant can access more data than another simply because they control a node — trust in the system collapses.
Cause 2: Confidentiality Concerns — Data on a Shared Ledger Is Never Truly Private
Even with Hyperledger Fabric's channel mechanism — the data isolation layer between participants — complete confidentiality of commercial data is not guaranteed. Transaction metadata, timing, and volumes can be analyzed even without direct payload access. For major carriers where routing information and cargo volumes are strategic assets, this was an unacceptable risk — not a technical concern to be mitigated, but a fundamental reason not to participate.
Cause 3: Lack of Interoperability with GSBN and Other Systems
Parallel to TradeLens, the Global Shipping Business Network (GSBN) was developing under a consortium of Asian ports and carriers. Instead of establishing a single industry standard, the market ended up with two incompatible ecosystems — forcing most participants to maintain integrations with both. This destroyed TradeLens's core value proposition: "a single source of truth for the entire supply chain".
Lessons from TradeLens for Enterprise Blockchain Projects
TradeLens is not a failure of blockchain technology. It is a failure of a specific governance model and adoption strategy. The technology worked — the pilot metrics confirmed that. The problem was who controlled the system and whose interests it actually served.
Specific architectural and organizational takeaways:
- Neutral ownership is not optional. Industry-wide platforms must be governed by an independent consortium or DAO structure — with no controlling participant that has competitive interests in the outcome.
- Testnet does not equal production. Pilot results from a controlled participant sample do not scale automatically. Validation on production data under real network conditions is a mandatory step before any scaling decision.
- Off-chain data requires a separate trust model. Blockchain records a hash — it does not verify the quality of incoming data. The oracle problem in logistics is not solved by smart contracts; it is solved by independent certification of IoT devices and data providers.
- Interoperability from day one. An industry platform without open standards and interoperability with competing systems is structurally destined for market fragmentation.
- ROI must be calculated for every participant, not just the initiator. TradeLens delivered obvious ROI for Maersk. What did MSC, CMA-CGM, or an independent customs broker receive in return? That answer needs to exist at project launch — not be discovered after the platform fails to attract adoption.
Where Blockchain in Logistics Lives Today
The closure of TradeLens does not signal the end of blockchain in supply chain. It signals a change in approach. Successful post-2022 projects share common traits: narrower scope, clearly defined bilateral or consortium models, and concrete measurable ROI for each participant.
CargoX — a blockchain platform for digital bills of lading — is integrated with customs systems in Egypt and Georgia and processes real import documentation. Unlike TradeLens, it does not attempt to unify the entire market. It solves a specific problem for a specific regulator, and that focus is what makes it viable.
For companies considering blockchain as a logistics optimization tool, the right starting point is not "which platform should we choose" but "what specific problem are we solving, and for which participants".
Technical Summary: What TradeLens Left Behind
TradeLens established three things that are now considered industry standards for enterprise blockchain:
1. Hyperledger Fabric is the right choice for permissioned enterprise use cases. Channels, chaincode, MSP (Membership Service Provider) — these mechanisms enable complex multi-party interactions with granular access control. The limitation: a centralized ordering service and significant operational complexity when managing a large number of participating organizations.
2. Smart contracts in logistics work — but require data standardization. Document automation through chaincode is real and delivers measurable impact. But only when all participants use compatible data formats. The absence of unified standards is the number-one technical blocker in any multi-organization blockchain deployment.
3. Blockchain is not a silver bullet for adoption. The technology solves the problem of data trust. It does not solve the problem of organizational trust, competitive interests, or regulatory compliance across multiple jurisdictions. Those problems are solved through governance, contracts, and regulatory engagement — blockchain only reduces transaction costs after a baseline level of trust has already been established.
If you are planning to implement blockchain in your business processes — the lessons from TradeLens deliver more practical value than most success stories. Failures teach with greater precision.
Why did Maersk and IBM choose Hyperledger Fabric over a public blockchain?
Hyperledger Fabric is a permissioned blockchain that enables role-based data access control. In shipping, this is critical: customs needs to see shipping documents, banks need financial transactions, and competing carriers should see nothing beyond their own data. Public blockchains (Ethereum, Bitcoin) cannot provide that granularity and do not scale to global shipping volumes. Additionally, Hyperledger Fabric supports channels — isolated sub-networks for bilateral interactions with their own ledger and chaincode.
What concrete results did the TradeLens pilot deliver?
Over one year of platform operation: 70–90% reduction in documentation costs (Mombasa → Rotterdam route dropped from $300 to $30–90); 40% reduction in US freight transit times; container status query steps reduced from 5–10 to 1–2. The flower delivery pilot showed that a single shipment which previously generated 200+ communications and a 25-centimeter paper stack could be fully automated through a single smart contract.
Why did TradeLens shut down if the technology was working?
The reason was not technical — it was governance. The platform designed to unite the entire container shipping market was controlled by Maersk, one of the largest competitors of its potential participants. MSC, CMA-CGM, and other major carriers refused to share operational data on a platform owned by their direct rival. Without broad participation, the network effect never materialized — and the platform failed to reach the commercial scale needed for viability.
What is the correct governance model for industry-wide blockchain?
Two viable options: (1) an independent consortium with equal representation from all major players and no controlling shareholder with competitive interests — analogous to the SWIFT model in banking; (2) fully decentralized on-chain governance with no single owner, where rules are encoded in smart contracts rather than corporate agreements. Any model where one competitor controls the platform will face the same adoption problem TradeLens faced.
Are there successful blockchain projects in logistics after TradeLens closed?
Yes. Successful post-TradeLens projects share a narrower scope and clear governance: CargoX (digital bills of lading, integrated with customs systems in Egypt and Georgia), We.Trade (trade finance for SMEs, banking consortium), and various port community systems with blockchain layers for document verification. All of them solve a specific problem for specific participants — and none attempt to replace the entire industry at once.
What is the oracle problem in the context of TradeLens?
The oracle problem is a fundamental challenge for blockchain systems: the blockchain can guarantee the immutability of recorded data, but it cannot verify the quality of data at the point of entry. In TradeLens, IoT sensor data (cargo temperature, GPS coordinates) was not stored directly on the blockchain — only its hash was. If a sensor or oracle transmitted incorrect data, the blockchain would faithfully record it as "verified". TradeLens addressed this through a network of trusted data providers — which effectively meant centralizing a critical component of a system built on the premise of decentralization.
