Rail has no shortage of technologies that could improve capacity, reduce energy consumption, and increase service frequency. What it has is a structural inability to adopt them at the pace the industry needs. The question worth asking is not why the technologies are slow, but rather why the decisions are.
About the Author (and article):
Alberto Mandler is CEO and Co-founder of DirectTrainS, an Israeli deep-tech company developing operational concepts for dynamic train formation on standard rail infrastructure.
This article is part of a series of guest-authored opinion articles. By publishing them, RailFreight.com hopes to engage the rail freight industry in a discussion on how to move the sector forward.
In my conversations with rail operators, procurement teams, and infrastructure managers, I encounter the same dynamic repeatedly. A technology is presented that demonstrably increases line capacity. The data supports it. The engineering case is clear. The commercial return, modelled conservatively, is positive. And yet the decision takes years, consumes enormous organisational energy, and frequently ends without a commitment — or with a commitment to a further study.
The instinct is to interpret this as conservatism, or risk aversion, or bureaucratic inertia. I do not think that is the right analysis. Rail decision-makers are not irrational. They are responding rationally to a specific set of structural incentives – and any technology that fails to engage with those incentives will face the same resistance, regardless of its technical merits.
The asymmetry that shapes every decision
The personal economics of a technology decision in a safety-critical industry are not symmetric. If a decision-maker approves a technology programme and it succeeds, they have done their job. Their organisation gains the projected benefits; they receive modest recognition. If the same decision-maker approves the same programme and it fails — or underperforms, or encounters unexpected regulatory complexity, or requires redeployment of assets that creates operational disruption — the consequences are disproportionate: regulatory inquiry, reputational damage, and career exposure that may be permanent.
This asymmetry is not unique to rail, but it is particularly pronounced in an industry where failure carries safety implications that extend into the public domain, and where regulatory scrutiny is continuous rather than episodic. A rational actor facing this calculus will not be moved by a compelling business case alone. They will seek certainty at a level that the business case cannot provide, because the downside they are managing is personal, not just institutional.
The result is a systematic tendency to defer, to require additional validation, to extend pilot phases, and to insist on precedent. None of this is irrational. All of it is explicable by the incentive structure in which these decisions are made.
The programme commitment trap
The second structural factor compounds the first. Most rail technology innovations require what might be called a ‘programme commitment’ before they deliver any return. New signalling standards require fleet-wide retrofit before the capacity benefit is realised. New coupling systems require compatible equipment at both ends of every intended coupling before the operational logic changes. New scheduling protocols require universal dispatcher training before the timetable benefits flow through.
This structure means that the decision to adopt a technology is not a decision to try just something – it is a decision to fund and manage a multi-year programme whose return is contingent on completion, not on progress. The investment precedes the anticipated benefit by years. And the personal exposure of the decision-maker who approved the programme is present throughout that period, before a single commercial return has materialised.
Consider Positive Train Control in the United States. Mandated by Congress following a fatal collision in 2008, with federal coordination, political support, and the full weight of regulatory obligation behind it. Full implementation across Class I railroads was not achieved until 2020. Twelve years. More than fifteen billion dollars across the industry. Every one of those twelve years, decision-makers across six major railroads managed the personal exposure of an incomplete programme. That is the structure of a programme commitment.
ETCS in Europe tells a similar story. Thirty years after development began, fewer than one in five rolling stock units carries onboard equipment. The technology works where deployed. The barrier is not technical, it is structural — the coordination burden, the investment horizon, the regulatory complexity, and the personal exposure that accumulates over years of incomplete deployment.
What containerisation got right
The most significant technology transformation in rail freight history was not a signalling upgrade or a new traction system. It was the standardised shipping container. Introduced commercially in 1956, it was eventually adopted universally, but not because the rail industry chose innovation proactively. Containerisation succeeded in rail because the competitive threat from trucking, enabled by expanding highway networks, made the alternative untenable. Rail operators did not adopt containers to capture opportunity. They adopted them to survive.
The lesson is not that rail only changes under existential pressure, though that is often true. The lesson is about what kind of adoption structure allowed containerisation to eventually succeed where other technologies stalled. A single operator could adopt containers on a single route, with a single shipper, without any dependency on what other operators did. The value was realised at the unit level. There was no programme commitment, no requirement for industry-wide coordination before the first return, and no regulatory certification path that blocked deployment. Each step funded the next.
A different kind of technology question
This analysis suggests that the industry’s adoption gap is not primarily a technology readiness problem. It is a decision structure problem. Technologies that require programme commitment before delivering return will continue to face the same resistance they always have. This is not because the people making the decisions are wrong, but because the structure of those decisions makes resistance rational.
The question that follows is whether a different kind of technology is possible: one that delivers measurable benefit from the first unit deployed, without dependency on broader network adoption. One whose failure mode does not affect the safety of the infrastructure it operates within. One that sits within existing regulatory frameworks without requiring new approvals. One that can be decided at an operational level — a route trial, a corridor pilot — rather than requiring programme board sponsorship and multi-year budget commitment.
If such a technology existed, it would change the personal calculus of the decision-maker. The downside exposure of approving a two-consist trial on one corridor is categorically different from the downside exposure of approving a fleet-wide programme. The former is an operational decision, whereas the latter is a career commitment. The structural incentives that currently produce resistance would, in that scenario, produce something closer to curiosity.
The harder conversation
The rail industry is having necessary and important conversations about ERTMS, moving block, and digital signalling. Those conversations should continue. The long-term capacity potential of a fully digitalised network is real, and the investment required to get there is worth making.
Still, the industry is not yet having an equally serious conversation about the decision structure that governs technology adoption — the incentive asymmetries, the programme commitment requirements, the regulatory coordination burdens that systematically advantage the status quo over change. Until that conversation happens, technologies will continue to move at the speed of the least-comfortable participant in the decision process, regardless of their merit.
The capacity problem is urgent. The technology options are improving. The limiting factor, increasingly, is neither. It is the structure of the decision itself — and that is something the industry has the power to change without waiting for any hardware to upgrade.
What would it mean for rail’s technology adoption pipeline if the minimum deployable unit were a pilot rather than a programme? Is the industry ready to design for that – or will procurement continue to treat every new technology as if it requires a systemic commitment before it can demonstrate its first unit of value?