Maritime Barrier Planning
PHAROS turns a coastline and a budget of buoys into a defensible sensor barrier, then keeps that barrier honest for its whole service life.
Self-hosted and air-gapped. Vendor neutral. Built by Cortex Prime Technologies.
The gap
You can buy software that watches buoys you already placed, software that positions a target inside an array you already built, and command systems that fuse sensors somebody else sited. Nobody sells the decision that comes first: where the sensors go.
Platform
Give it a sector and a number of buoys. It returns a verdict, a gap list, a proposed layout and the cost to close. Ask it the other way and it sizes the fleet for a required detection probability. Every assumption is on screen and contestable.
State, alerts and detections from the deployed network, delivered into the command system you already operate. PHAROS is a sensor source, not a second screen for an operator who is already busy.
Charge cycles and cell health project each node's end of life. The platform names the date and the kilometre mark where the barrier will fail, then routes the boat against port distance, sea-state windows and how critical each gap would be.
Every placement decision recorded with author, time and rationale, immutably. Export is role-gated and logged, because the true position of a detection barrier is the most sensitive object in the system. Generates the AtoN authorisation dossier where one is required.
What it computes
A depth profile is traced along every azimuth. The beam is clipped where it meets the seabed, given transducer depth and vertical beamwidth. The result is a polygon, not the circle on the brochure.
For a defined target crossing at a defined angle, the cumulative probability across every overlapping sensor on its track. Binary covered or not covered is what the market reports, and it overstates reality.
A sonar contact only means something when the radar sees nothing at the same coordinate. So a sonar node outside radar coverage raises no usable alarm. Protected water is the intersection, never the union.
Three nodes within range but nearly collinear give a poor position fix. Dilution of precision is checked across the chain, and layouts are rejected on geometry even when every distance is legal.
Effective spacing is nominal spacing minus both watch-circle radii, derived from charted depth and mooring scope. Pack too tight and the field becomes a charted obstruction that must be marked, which ends any silent regime.
Satellite airtime for a large fleet is not affordable, and video cannot cross it at all. Nodes must process at the edge and reach a gateway within a bounded number of hops. That constrains where a node can physically go.
Radar cuts the node count but not always the bill, and it is blind alongside a quay. Video sees everything inshore and bankrupts you offshore. PHAROS finds the boundary, per square kilometre covered, and prices both sides of it.
Every one of these compounds. A sensor whose real footprint is 28 % smaller than the datasheet, sitting where the geometry is weak, with a battery three months from the cliff, inside a chain whose gaps nobody measured, produces a coast that looks defended on a map and is not. PHAROS exists to make that gap visible while it is still a line item and not an incident.
The scoping question
Not by a factor of technology. By one decision that usually has not been made yet: what counts as a place that needs eyes.
Ports, marinas, estuary mouths, islands. The places where a small craft hides among legitimate traffic and where radar is useless.
Open frontage added to the pockets. Cheaper per kilometre, far more kilometres, and the point where sensor choice starts to dominate the bill.
Layered coverage everywhere, to a stated detection probability. Defensible on paper and usually undone by something other than capex.
Each of these is a legitimate answer, and each is a different programme. PHAROS prices all of them against your actual sensor payloads, your seabed and your maintenance capacity, so the choice is made with arithmetic rather than instinct. The constraint that decides it is rarely the purchase price: a large moored fleet is measured in boat-days per year, and that number ends more programmes than capex ever does.
Deployment
These decisions were taken before the first line of code, because none of them can be retrofitted afterwards.
Ships as an appliance on the customer's own infrastructure. Bathymetry and chart data are imported, never queried at runtime. A production install has no public hostname at all.
Separate deployments, not row-level filtering in a shared database. One customer's theatre does not physically exist inside another's.
Whether a node broadcasts is a first-class property, enforced on every export path and outbound interface. A silent node cannot leak through a report written for the public ones.
Sensors are parametric specifications, not code. A new payload from a new supplier is a form to fill in. You are not buying a planning tool welded to one manufacturer's catalogue.
Pilot programme
A pilot starts with one stretch of coast, your actual sensor payloads and the buoy count you can afford. Inside two weeks you get a costed layout, a modelled detection probability and the list of things your current plan would have missed.
Evaluation access is granted on request and carries synthetic data only. Operational deployments run on your infrastructure, disconnected.