Most organisations have dashboards. They have alerts. They have incident response plans sitting in a shared drive that was last updated in 2023. What they do not have is a way to answer this question before the incident happens:
Given what is coming in right now, what is the most likely operational outcome – and how long do I have to change it?
That is the question the OCTA model was built to answer.
The Problem With Reaction
Reaction is not the same as readiness. When a signal arrives – a navigation anomaly, a supplier failure, a system degradation, an infrastructure event – most organisations begin a process: escalate, assess, convene, decide. By the time a decision is made, the intervention window has often already closed. The cost is not the incident itself. The cost is the gap between when the information arrived and when the organisation acted on it. That gap is not random. It follows patterns. And patterns can be modelled.
What OCTA Does
The OCTA model takes incoming information from operational systems, external data sources or environmental signals and maps it against a dependency structure to calculate two things:
- The most probable outcome if the current trajectory continues.
- The intervention window: how long the organisation has before that outcome becomes difficult or impossible to reverse.
It does not predict the future. It predicts the consequences of inaction given what is already known. The model uses a Dependency Cascade Index to measure how tightly coupled an organisation’s critical functions are to the incoming signal. The more coupled, the faster the cascade. The faster the cascade, the shorter the window. The output is not an alert. It is a probability-weighted outcome with a time dimension attached.
Aviation
The situation
An aircraft’s navigation systems receive position data continuously. When that data is degraded or spoofed, the flight management system keeps making decisions, and the crew keeps flying, on information that is no longer true. The aircraft does not stop. It continues with full confidence on a heading that is wrong.
What follows
- Two aircraft converging on the same coordinate, each believing the other is elsewhere.
- A heading correction that costs nothing at minute 4 becomes a missed approach at minute 12.
- A missed approach becomes a diversion – grounding the aircraft, breaking the crew schedule, rippling across every flight behind it.
- ATC, dispatch and maintenance are all informed after the cascade has already run.
What OCTA changes
A single diversion costs between £80,000 and £200,000 before the schedule cascade is counted. OCTA calculates the intervention window in real time so the decision to act reaches the right people while the cheap option is still on the table.
Oil and Gas
The situation
Offshore platforms, pipeline networks and tanker fleets operate continuously on timing and position data they cannot afford to question. When that data degrades, nothing stops immediately. The platform holds station. The pipeline keeps reporting. The tanker keeps broadcasting. All of them, on data that is silently wrong, for as long as it takes someone to notice.
What follows
- A platform’s dynamic positioning system drifts off station into a restricted zone.
- Pipeline integrity timestamps drift until the monitoring record is no longer valid.
- A tanker broadcasts a false position into a congested shipping lane.
- The operations centre reads normal across every screen until something physical happens.
What OCTA changes
Platform downtime runs at over $1 million per day before environmental costs are counted. A safety-triggered shutdown takes weeks to reverse. OCTA maps which threshold gets crossed first and how long before it happens, so the team can act while the platform is still on station.
Banks
Nine of the UK’s biggest banks. Two years. That is not bad luck. That is a pattern nobody measured until customers were calling food banks.
803
hours offline
158
incidents
33
days lost
Banking infrastructure does not fail at the point anyone expects it to. It degrades across layers that are each measuring their own picture, and nobody is measuring the combined one. When Barclays’ mainframe started degrading – not failing, degrading – every dashboard still showed green. By the time anyone called it critical, 56% of online payments had already failed. Customers locked out for three days. Up to £12.5 million in compensation. The window where a controlled fix was still possible opened and closed while nobody was watching it.
When the ECB’s T2 settlement system failed, over three trillion euros of daily payments across the eurozone stopped for seven hours. When CrowdStrike pushed one bad software update, JPMorgan, Bank of America and trading desks from Hong Kong to Dubai went dark at the same time. One software layer cascading into everything that depended on it. When fifteen Danish banks went down simultaneously through a shared infrastructure failure, it was not fifteen separate problems. It was one combined picture nobody had measured.
What OCTA changes
OCTA watches the layers degrading together, calculates that combined picture in real time, and produces a number: how long before the window closes. Not a warning that the system is already down. A figure telling you how long you still have to act.
Pharma
The situation
A pharmaceutical supply chain is an unbroken chain of dependencies from a manufacturing facility to a patient. Every link – cold storage, logistics handoffs, batch records, distribution centres – depends on the one before it being accurate. When one link reports incorrectly, the chain does not stop. It keeps moving, carrying a problem it does not know it has.
What follows
- Affected batches are distributed across multiple markets before anyone knows which batches are affected.
- The quarantine window closes. A voluntary action at hour 6 becomes a mandatory one at hour 24.
- Production lines halt while the audit trail is reconstructed. Hospitals run short.
- Patients receive compromised product.
What OCTA changes
A voluntary recall averages $10 million. A mandated one runs two to three times that. Both figures are secondary to what caused them. OCTA traces the chain in real time and tells the quality team exactly where they are in the window while the option to contain it quietly is still open.
Automotive
The situation
A just-in-time production line has no slack. Every component arrives at the point it is needed and not before. When a supplier fails, the line does not slow down. It runs until it cannot, and then it stops. The damage is not contained to the factory floor. It is company-wide from the moment the line stops.
What follows
- The tier-two dependency behind it was already failing before the first alert fired.
- Rerouting options available at hour 2 are gone by hour 6. Air freight replaces them at four times the cost.
- The line stops. Every hour it is stopped costs between $1 million and $2 million in lost production.
- Customer penalty clauses activate. Sales teams lose delivery commitments. Finance revises the quarter.
Every hour the line stops
$1–2M
in lost production – before penalty clauses and downstream commercial damage
What OCTA changes
One delay in one part of the operation does not stay there. It moves through procurement, production, logistics, sales and finance, hitting each one in sequence. OCTA is the umbrella across that whole cascade. It calculates the sequence before it runs, attaches a time to each stage, and gives procurement, operations and finance the same picture simultaneously. One model. One view. The decision lands while the cheap option is still on the table.
Private Equity
The situation
A PE firm holding a portfolio company in any of these sectors is holding its operational risk whether it has mapped it or not. Financial due diligence tells you what the company has done. It does not tell you what the company’s operations will do the first time something goes wrong. That gap lives between acquisition close and the moment the cascade starts, and it is invisible until it is not.
What follows
- Customer contracts break. Revenue stops. Counterparties invoke force majeure.
- Management time disappears into the incident. Everything else in the business slows.
- The valuation impact is not the direct cost. It is the sum of every downstream consequence.
- At exit, an unmapped dependency structure is a discount the buyer extracts before signing.
A portfolio company with a mapped and monitored dependency structure is a different asset than one without it. That difference is negotiated in the room.
What OCTA changes
OCTA applied at due diligence maps the dependency structure before the acquisition closes. Applied during the hold period, it monitors for the incoming signals that predict cascade risk across the portfolio before they become incidents. Applied at exit, it is evidence that operational continuity has been actively managed – a valuation argument at the negotiating table.
One Model Across the Whole Company
Every sector above has one thing in common: the damage did not stay where it started. A delayed component is not a supply chain problem. It is a production problem, then a revenue problem, then a customer relationship problem, then a finance problem. A payment system going dark is not an IT problem. It is a customer problem, a regulatory problem, a reputational problem. The cascade does not respect department boundaries. It moves through an organisation until it runs out of things to affect.
OCTA is the umbrella across that whole picture. It does not sit in one department or monitor one system. It maps the dependency structure of the company as a whole, watches for the incoming signals that predict cascade risk, and produces a single output: how long before the window closes across every function that will be affected. That is not a dashboard. It is not an alert. It is a prediction, company-wide, with a time attached to it.
The working paper on operational continuity in GNSS-denied environments, which introduces the model and its mathematical basis, is available here: jameslarsson.pro/research-octa
