Piranha USV, strike and surveillance from one hull
A small unmanned surface vessel carrying a marine radar and a stabilised EO/IR payload. The same hull works as a reconnaissance craft, building a radar and camera picture of the coast and shipping lanes, and as a one-way strike craft with a 50 kg+ class warhead. Top speed is 65 km/h.
The hull and onboard hardware are in build now, with sea trials from the end of 2026. Navigation and control follow the core Apollyon approach: physical references the vessel senses for itself, no dependence on GNSS, and autonomy that runs on board.
Primary engineering baseline
Piranha is not a cheap boat. A marine radar and a stabilised EO/IR gimbal are the largest items in its ₹1.5–2.5 crore estimated cost; the hull, engine and waterjet are a smaller share. That buys a craft that can watch a stretch of coast for days in peacetime and, in the strike configuration, find and hit a target on its own sensors. It is still a small fraction of the cost of a crewed patrol vessel or an anti-ship missile.
Internal architecture & CAD layout
Primary use cases
Piranha works in littoral waters, island chains and approaches to ports, where a crewed boat would be at risk. Every craft carries the radar and EO/IR payload; strike craft add the warhead.
| Mission | Operational problem | How Piranha is used |
|---|---|---|
| Coastal reconnaissance | Keeping a continuous picture of shipping lanes, coastline and island territory without tying up crewed vessels. | Patrols with the marine radar and EO/IR gimbal running and relays tracks and imagery to shore. |
| Littoral strike | Hostile small combatants, landing craft and supply boats operating close to the coast. | A strike-configured craft finds the target on its own radar and camera, runs in at up to 65 km/h and detonates at the waterline. |
| Coordinated strike | A single small craft is easy for a defended vessel to track and engage. | Several craft coordinate their arrival over a mesh link and approach from different bearings, splitting the defender's attention. |
| Port & anchorage denial | Defended harbours and anchorages under GNSS jamming. | Navigates by shoreline scene matching and targets moored vessels and jetty infrastructure. |
Navigation, control and warhead
A · Non-satellite sea-surface navigation
Satellite denial over open water presents unique challenges: unlike land terrain, the open sea has no static topographic features for elevation correlation. The Piranha USV employs a layered navigation stack built specifically for marine environments:
- CRPA Anti-Jam GNSS: Indigenous multi-element antenna suppresses coastal jammers by >45 dB, maintaining NavIC/GPS lock during initial transit.
- Hydro-Inertial Dead Reckoning: Tactical IMUs fused with a Doppler Velocity Log (DVL). In coastal water the DVL locks onto the seabed and measures true speed over ground, which removes current drift; beyond its bottom-lock depth it measures speed through the water, and the current becomes an estimated state.
- Coastline DSMAC: As the vessel nears target archipelagos, ports, or coastlines, high-resolution optical cameras match shoreline profiles, headlands, and navigation marks against pre-loaded commercial satellite maps.
- Optical Waterline Tracking: In the terminal run, the stabilised EO/IR gimbal locks onto the target's hull silhouette at the waterline, which chaff and most decoys do not reproduce.
B · Control: the same method, a marine controller
Piranha is controlled the way the aircraft are, with the physics changed from air to water. A guidance layer turns the route into the heading and speed to hold, a controller holds them, and the waterjets carry out the commands. As on every Apollyon platform, the controller's settings come from a model of the vehicle identified from real runs, not from textbook assumptions, and envelope limits keep demand inside what the hull and jets can actually deliver.
The hard part is the equivalent of flying near the edge of the envelope. A fast hull behaves very differently at low speed, while climbing onto the plane and at full sprint, so the control settings are scheduled across those regimes rather than fixed. The sea adds its own problems. The controller has to ignore the rocking of individual waves and follow only the real motion. It has to hold running trim so the hull stays stable at speed, and it has to keep station on the track despite current and wind pushing the vessel sideways.
- Speed follows the sea: as the sea state rises, the vessel limits its own speed to stay inside the structural limits of the hull.
- Sharing the water: in transit and on patrol it keeps clear of other traffic using its marine radar and EO/IR, and follows the rules of the road.
- When the link drops: it falls back to a pre-set safe behaviour. The decision to strike always stays with a human commander.
The hull and hardware are still in build, so these settings will be worked out in sea trials, through the same loop the aircraft use: run the vessel, compare the logs with the simulation, correct the model and the settings, and clear each change in hardware-in-the-loop testing before it goes back on the water. One method, many platforms →
C · Warhead
The strike configuration carries a 50 kg+ class high-explosive warhead in the bow bay. It is sized for small combatants, landing craft, moored vessels and jetty infrastructure, not for major warships.
- Waterline detonation: the craft is guided onto the target's hull at the waterline, where damage lets water in.
- Redundant fuzing: contact and delayed-action fuzes, so a glancing hit still detonates.
- Human decision: the decision to strike stays with a human commander.
One hull, two customers
One hull serves both jobs. Every Piranha carries the marine radar and the EO/IR gimbal. The reconnaissance configuration uses them to watch the coast and shipping lanes and comes home; the strike configuration adds the 50 kg+ warhead and uses the same sensors to find and hit the target. Coast guards, port authorities and littoral security forces get a persistent sensor picture without crewing a boat, and the peacetime fleet keeps the boatyard tooling and the crews who operate it in practice.
Hull and onboard hardware in development through 2026; sea trials with the radar and EO/IR payload from the end of 2026; serial production from 2028. The hull, sensors and navigation package are common to both configurations.
Subsystems powering this vessel
Robust flight control · the method
Piranha runs its own marine autopilot. What it inherits from the aircraft programmes is the method: instrumented trials, a model identified from the logs, and hardware-in-the-loop clearance before each configuration goes on the water.
GNSS-denied navigation
Anti-jam CRPA NavIC/GPS, coastline scene matching and hydro-inertial dead reckoning for long sea transits.
Edge compute
Runs waterline tracking, collision avoidance, the sea-state speed governor and mesh coordination on the mission computer.