Medjay Technologies | From Sensing to Decision Superiority

// OUR PLATFORMS

Decision Intelligence Platforms

FFDCS

AI-Based Fire Fighting & Damage Control System

AI-Based Fire Fighting & Damage Control System

The Operational Reality

Fire. Flooding. Simultaneously.
Four minutes to make the right call.

Fire fighting and damage control aboard a naval vessel is among the most time-critical, cognitively demanding operations. When an incident occurs, the Commanding Officer and damage control team face an immediate cascade of simultaneous variables — each one capable of determining whether the ship survives.

"The challenge is not courage. It is clarity — the ability to make the right decision while processing numerous inputs, under extreme stress, in a short window."

— Damage Control doctrine

In those first critical minutes, the team must simultaneously assess where fire is spreading, where flooding is ingressing, which compartments require boundary cooling, which escape routes remain viable — and critically, what the evolving incident means for the ship's ability to Float, Move and Fight.

🚢

Float

Flooding threatens buoyancy and stability. Every second of ingress shifts the GZ curve.

⚙️

Move

Engine room fire eliminates propulsion. Loss of mobility in a threat environment is critical.

🎯

Fight

Damage to weapons, power, and command capability degrades operational effectiveness.

14+

Simultaneous alarms during a combined fire & flood incident

<4 min

Window before a contained fire reaches adjacent fuel systems

Critical decisions required in the first 60 seconds

∞

Variables interacting simultaneously — fire, flood, stability, manpower

The Gap

Traditional damage control
was not built for this complexity.

Until now, damage control teams have relied on paper DCS manuals, radio coordination, trained instinct, and manual stability calculations — tools designed for a simpler operational environment. These methods place the entire cognitive burden on the officer in command, at the moment of maximum stress.

No real-time propagation modelling. Fire and flood spread is estimated from training and memory — not computed from live sensor data and ship geometry.

No dynamic stability computation. GZ and heel calculations are manual, time-consuming, and impossible to perform accurately during an active incident.

No ranked decision support. The officer must recall SOPs, assess priorities, and sequence actions from memory while managing communications and personnel.

No crew location intelligence. Rounds monitoring and personnel accountability rely on manual reporting — creating dangerous blind spots.

No post-incident learning loop. Debriefs rely on subjective recall. Actions taken too late, too early, or incorrectly are not systematically captured.

The Solution

FFDCS closes the decision gap.
Superior situational awareness. AI-enabled command.

The moment an incident is detected, FFDCS integrates all sensor inputs, highlights affected and adjacent compartments, generates AI-ranked actions, runs propagation simulation in parallel, computes live stability — and broadcasts to the ship. All simultaneously. In seconds.

9 sensor types — one unified platform

Fire detection

🔥

AFDS

Auto Fire Detection System

TRIGGERED

▾ Sub-sensors

Smoke detectors14 zones

Fire detectorsAER triggered

IR sensors842°C

🌡️

Firemain Pressure

Fire suppression pressure

8.2 bar — nominal

⚠️

NBC Sensors

Nuclear · Biological · Chemical

Clear

Flooding & damage

🌊

AFAS

Auto Flood Alarm System

Ingress detected

💧

Flooding parameters

Inflow: 1.80 m³/s · Cd: 0.60 · Compartment: AER · Side: Port

Rate mode active

Situational awareness

📷

CCTV

Visual surveillance — all decks

Live feed — AER

📡

RFID

Rounds monitoring — crew location

Last round: T−04:12

🎥

Body Worn Camera

First responder live feed

DRT feed active

Dual-screen response initiated simultaneously

Action screen2D deck plan

Compartment map

BRIDGE

OPS RM

FWD MESS

FWD STORE

AMMO

GEAR RM

AER ●

AUX RM

IncidentBoundary cooling

AI priority actions

P1Activate fixed CO₂ — AER

P1Boundary cooling — Gear Rm & Aux Rm

P2Close WT doors — Frame 42

P3Evacuate — Port passageway

Info screenSimulation running

Fire propagation

AER (origin)

—

Gear room

—

Aux room

—

Flood dynamics

Inflow rate

—

Fill level

—

RFID — crew last seen

■ PO Sharma — AER corridor — T−02:14

■ Lt. Mehta — Gear Room — T−05:41

■ CPO Rajan — Bridge — T−00:32

Built-in stability module — NES 106 & GZ computations

3.2°

Heel (live)

0.4°

Trim (live)

0.51 m

GZ (live)

0.000 m

GM eff — damaged

GZ CURVE — INTACT VS DAMAGED (LIVE)

Ship-specific 3D model · NES 106 compliant · Stability recomputed as flooding changes weight distribution in real time

Real-time AI analyser — every action scored as it happens

Correct

CO₂ system activated in Engine Room — within SOP window

T+00:42 · 8s before SOP deadline · Optimal

Correct

Boundary cooling initiated on adjacent compartments

T+01:14 · Consistent with past incident data

Taken late

Watertight doors — Frame 42 closed 34s after recommended window

T+02:18 · SOP window: T+01:44 · Delay logged

Wrong step

Ventilation engaged — increases fire spread risk in this scenario

T+01:55 · Contraindicated by AI model · Correlation: 87%

Runs in real time · SOP logic + trained on past incident data · Full debrief report auto-generated on resolution

Auto broadcast — integrated with ship announcement system

📢

AUTO ANNOUNCEMENT — GENERATED & TRANSMITTED BY FFDCS

"Fire fire fire — Engine Room — Compartment AER — Port side. Damage Control Team close up. All personnel evacuate via Port passageway. This is not an exercise."

Triggered at T+00:08 · Repeats every 30s until acknowledged by Officer of Watch

9 sensor types — one unified platform

2D deck plan · escape routes · equipment

Fire & flood dynamics simulation

Ship-specific 3D digital twin

NES 106 & GZ stability module

Real-time AI decision analyser

RFID rounds monitoring · crew location

Auto ship broadcast integration

1 of 5

The Outcome

Faster decisions. Protected crew.
Ships that Float, Move and Fight.

FFDCS does not replace the officer in command — it amplifies their capability. By converting sensor data into situational awareness, simulation into foresight, and SOPs into ranked action, FFDCS gives commanders decision superiority in the most demanding moments of naval operations.

⚡

Faster Response

AI-ranked priority actions delivered in seconds. No manual SOP lookup. No memory under stress.

🧠

Reduced Cognitive Load

The system manages complexity. The officer manages command. Each focused on what only they can do.

🛡️

Crew Safety

RFID tracking, escape routes, body-worn feeds and auto-broadcast keep every crew member accountable.

📈

Continuous Improvement

Every incident is analysed in real time. Every action scored. Every exercise becomes a data point.

🏆

IDEX Recognised

Winner of the IDEX Challenge — recognised internationally for innovation in AI-based damage control.

🔩

Ship-Specific

Every deployment configured with the exact digital twin of that vessel — not a generic template.

// REQUEST A DEMONSTRATION

See FFDCS deployed on a ship-specific scenario. Our team will configure a demonstration using a vessel class relevant to your requirements.

PRAVA

AI-Powered Language, Vision & Decision Support Platform

The Operational Reality

Every system sees a piece.
None of them see the picture.

Across defence, security and industrial environments, cameras, alarms, logs, operational doctrines, sensor streams and operator inputs exist in isolated silos. Operators are expected to synthesise all of this under time pressure — often during the most critical moments of an operation.

"Conventional systems provide monitoring. The gap is decision support — what to do, when, and why."

— PRAVA Design Philosophy

📡

Too Many Silos

Cameras, sensors, logs and documents generate data independently with no unified intelligence layer.

🧠

Human Overload

Operators must synthesise fragmented inputs under extreme time pressure at the moment of crisis.

☁️

Cloud Dependency

Most AI platforms require cloud — unsuitable for defence, sensitive or low-connectivity deployments.

Domain-specific solution packages on one common AI core

Cloud dependency — fully offline, edge-first architecture

15+

Use cases across defence, security and industrial environments

Unified platform core — build once, deploy across all domains

The Gap

Current systems generate alerts.
None recommend actions.

The operational gap is not detection — it is the transition from detection to decision. Most platforms stop at the alarm. PRAVA starts there.

Too many standalone subsystems, too little fusion. Each domain has its own tool — none produce a unified operational picture.

Heavy dependence on human interpretation at the moment of crisis. Operators correlate fragmented inputs manually under time pressure with no AI assistance.

Cloud-first AI approaches unsuitable for sensitive deployments. Latency, confidentiality and resilience requirements make cloud-dependent platforms unacceptable for defence.

No comparison of live events against operational doctrines. No system cross-references what is happening now with what doctrine says should happen next.

Weak transition from detection to decision. Many systems generate alerts but few recommend actions — the hardest cognitive work is left entirely to the operator.

The Solution

PRAVA closes the decision gap.
One platform. Four domains. Offline-first.

PRAVA fuses data from cameras, sensors, documents and operators — converting it into actionable intelligence. No cloud. No fragmentation. One coherent decision support layer across all your critical environments.

INPUTS

📡 Sensors & IoT

📷 Vision / Cameras

📄 Documents & SOPs

👤 Operator Inputs

🔍 OSINT Feeds

PRAVA

AI CORE

OUTPUTS

Situational Awareness

Ranked Decisions

Evidence Packages

Alert & Action Console

Doctrine Compliance

🔒

Offline-First

Air-gapped, edge deployments. No cloud dependency. Built for sensitive environments.

🔍

Evidence-First

Every output shows why the system reached its conclusion — not a black box.

👤

Human-in-Loop

Decision support, not automation. The operator commands. PRAVA informs.

🧩

Modular Domains

One core. Four domain packs. Build once, deploy across defence, security, industrial and enterprise.

PRAVA AI — OFFLINE · SECURE MODE

● EDGE DEPLOYMENT

PRAVA // READYOperational in offline-secure mode. Zero cloud dependency. Try: threat, drone, intrusion, PPE, fire, document

The Four Domains

One core. Four mission packages.

A defence customer should not need to interpret industrial use cases. PRAVA is packaged domain-first — same AI core, purpose-built for each environment.

PRAVA DEFENCE

Tactical Awareness & Mission Support

AI-enabled decision support for defence and armed forces. Fuses sensor feeds, camera inputs and operational intelligence into a unified tactical picture — without cloud dependency.

Drone & swarm detection — UAV classification and graded threat alerts

Tactical zone monitoring — personnel & vehicle tracking in defined areas

Sensor fusion for threat assessment — camera, acoustic, RF and operator inputs

Autonomous anomaly spotting — loitering, unusual movement, rapid direction change

Command decision support — ranked response options with evidence packages

ISR augmentation — intelligence, surveillance and reconnaissance support

PRAVA DEFENCE // THREAT ASSESSMENT

HIGH

Drone swarm — Sector 7 North

3 UAVs · Bearing 042° · Range 800m · Closing

MED

Unauthorised vehicle — Zone Alpha

Loitering 14 min · No IFF · Monitor

INFO

Acoustic anomaly — Perimeter East

Matched movement pattern · Correlating cameras

RECOMMENDED ACTION

Activate counter-UAV protocol · Alert sector commander · Initiate IFF sweep

PRAVA HOMELAND SECURITY

Urban Surveillance & Security Operations

AI-powered surveillance intelligence for public safety, critical site protection and urban security. Real-time multi-camera correlation — no cloud, no latency.

Crowd density & people counting — occupancy, queue and density monitoring

Watchlist & facial match alerts — entry control against local databases

Suspicious behaviour — loitering, restricted zone, abandoned object detection

Perimeter intrusion detection — virtual geofenced zone breach alerts

Multi-camera event correlation — linked incident trails across adjacent feeds

PRAVA HOMELAND // LIVE ALERTS

ALERT

Perimeter breach — Gate 3 East

Unauthorised entry · Camera 7 · T+00:23

WATCH

Crowd density critical — Terminal B

Occupancy 94% · 3 cameras correlated

INFO

Abandoned object — Platform 2

Stationary 8 min · No owner detected

PRAVA INDUSTRIAL DSS

Safety, Compliance & Plant Intelligence

AI decision support for industrial safety, operational compliance and equipment intelligence. Fuses vision, sensor and document inputs into actionable plant intelligence — on-premise.

PPE compliance monitoring — hard hats, safety vests in industrial work areas

Fire & smoke detection — early incipient detection from video feeds

Leak detection — sensor fusion with visual confirmation for early anomaly detection

Restricted zone violation — human presence in unsafe or high-risk areas

Equipment anomaly detection — unusual machine states from image and vibration

Doctrine compliance engine — live actions vs documented procedures, deviations flagged

PRAVA INDUSTRIAL // SAFETY DASHBOARD

CRITICAL

PPE violation — Zone C Welding Bay

No hard hat detected · Alert supervisor

WARNING

Pressure anomaly — Compressor Unit 4

+18% above baseline · Vibration elevated

COMPLIANT

SOP audit — Shutdown Procedure 7B

All 12 steps completed · No deviations

PRAVA CORE

Secure On-Premise AI for Knowledge & Documents

For any organisation requiring on-premise AI with document intelligence, OCR and a secure knowledge assistant. No data leaves the building. No cloud. Full control.

Intelligent Document Search & Q&A — RAG-powered knowledge retrieval

Scanned Document Digitisation — OCR plus structured data extraction

Form & Report Data Extraction — automated intake from structured documents

Internal Knowledge Assistant — chat with your own documents, fully offline

Basic Object Detection — vision AI for access, compliance and monitoring

PRAVA CORE // KNOWLEDGE ASSISTANT

PRAVA // OFFLINESecure knowledge mode active. Your documents are indexed locally. What would you like to know?

What does our emergency shutdown procedure say about compressor isolation?

PRAVA // RESPONSEFrom Procedure ESD-7B, Section 4.2: Isolate compressor feed valve V-12 before initiating shutdown. Confirm pressure below 2 bar before step 3. Source: Page 14, para 3.

The Architecture

Edge-first. Modular. Air-gapped.
Built for environments where cloud is not an option.

Input Layer

IP CamerasThermalMicrophonesModbus SensorsDocumentsOperator Inputs

▼

Ingestion Layer

Video Stream HandlingSensor PollingEvent NormalisationDocument Ingestion

▼

Intelligence Layer

Vision AI ModelsRAG / Knowledge EngineRule EnginesLocal LLMPattern Analysis

▼

Decision Layer

Alert ScoringDoctrine ComparisonRecommended ActionsEvidence Packages

▼

Presentation Layer

Operator DashboardAlert ConsolePlayback & ReviewRole-Based AccessAudit Logs

// REQUEST A DEMONSTRATION

See PRAVA deployed in your domain. Our team will configure a demonstration tailored to your operational environment.

DRISHTI

Dynamic Risk Intelligence & Situational Threat Index

Dynamic Risk Intelligence & Situational Threat Index

The Intelligence Challenge

Decision-makers are drowning
in signals they cannot act on.

Geopolitical risk is rarely a surprise in hindsight. The signals were there — troop movements, rhetoric shifts, policy changes, economic stress. The failure is not a lack of data. It is the absence of a structured, calibrated system that converts fragmented signals into a clear, auditable picture of where a situation is heading.

"The challenge for leadership is not more information. It is knowing — with quantified confidence — what the probability of escalation is, and why it changed since yesterday."

— DRISHTI Design Philosophy

📰

Signal Overload

OSINT sources, sensor feeds and field reports flood in — no system to weight, fuse and rank them into a coherent picture.

🎲

Gut-Feel Assessments

Analyst judgements vary by individual, lack calibration, and cannot be audited or compared against historical accuracy.

⏱️

Latency to Decision

By the time intelligence reaches a briefing, the situation has moved. Leadership needs live probability — not retrospective reports.

Escalation states — S0 Stable through S4 Crisis to S5 Recovery

Active theatres monitored simultaneously — live and auditable

Mathematical models — Markov, Bayes, Shapley, Brier calibration

Questions DRISHTI answers for every decision-maker, every briefing

The Gap

Intelligence products tell you what happened.
Not what is next.

Traditional risk tools are built for analysts producing reports. DRISHTI is built for decision-makers who need to act. The gap is calibrated, live, auditable probability at executive tempo.

Point estimates without confidence bands. "70% probability" means nothing without the uncertainty range. DRISHTI shows both — always.

No audit trail for probability changes. When the assessment shifts overnight, there is no answer to: what changed, which signal moved the needle, and by how much?

Analyst subjectivity without calibration. Judgements are rarely scored against outcomes. DRISHTI uses Brier scoring to continuously calibrate the model against reality.

No driver attribution. Even when a situation changes, no system says which indicator was most responsible. Shapley attribution solves this.

Static briefings at fixed intervals. A morning report is stale by noon. Leadership needs live state probability — not a daily slide deck.

The Solution

DRISHTI answers three questions.
Every briefing. Every theatre. Live.

How likely is escalation?

Probability with confidence bands — not a point estimate. S0 through S5 state distribution, live and updated on every signal change.

Why has the likelihood changed?

Every probability movement is auditable. Which signal triggered it. Shapley attribution shows the contribution of each driver.

What should be watched next?

The system surfaces indicators most likely to drive the next state transition — so analysts and decision-makers know where to focus.

// DRISHTI LIVE · UKRAINE WAR THEATRE

S0 Stable

0.00%

S1 Fragile

0.03%

S2 Tense

0.57%

S3 Escalating ▲ NOW

39.32%

S4 Crisis • NEXT

60.08%

S5 Recovery

0.00%

ACIED INDICATORS

Troop movement

100%

Hostile rhetoric

73%

Remote violence

100%

Regulatory advisory

100%

Social unrest

80%

RECOMMENDED DECISION

Airspace Closure Advisory
Heavy artillery near Pokrovsk. Drone strikes on military positions. Missile barrage targeting Zaporizhzhia.

Bayesian update +25.1pp • Shapley: Troop movement primary driver

The Architecture

Three explicit layers.
Signals. Inference. Decision.

DRISHTI maintains strict separation between raw data, mathematical inference, and what leadership sees — keeping every layer auditable and every output explainable.

📡

Layer 1 — Signals (Raw Indicators)

Raw indicators ingested, classified, weighted and normalised before reaching the inference engine.

OSINT feedsSensor dataField reportsPolicy shiftsACLED eventsGDELTEconomic indicators

▼

🧮

Layer 2 — Inference Models

Four mathematical models working in concert to produce calibrated, attributable, auditable probability estimates.

Markov State Ladder

S0 through S5 escalation states with defined transition probabilities. Every state change is modelled, not guessed.

Bayesian Updates

Disciplined probability revision as new signals arrive. Prior beliefs updated with evidence. Every shift is logged.

Shapley Attribution

Identifies which indicator contributed most to a probability change — troop movement, rhetoric, or economic stress.

Brier Calibration

Continuous scoring of forecast accuracy against outcomes. The model is held accountable, not just confident.

▼

🎯

Layer 3 — Decision Layer

What leadership sees. Probability with confidence bands, driver attribution, next-watch indicators — at executive tempo.

State probability + confidence bandsFull audit trailShapley driver cardsNext-watch signalsMulti-theatre dashboardMap overlaysAI Insights brief

The Dashboard

Built for decision-makers.
Not for analysts writing reports.

One screen tells the complete story — current state, probability distribution, live event feed, indicator weights, map overlays and AI-generated brief. Everything auditable, nothing buried.

DRISHTI LIVE DASHBOARD

● LIVE • GDELT + USGS

MAP LAYERS • STRATEGIC POSTURE

⚠ IRAN THEATER — ELEVATED

🌍

5 Active Theatres

Ukraine, Iran-Israel, Sahel, Taiwan Strait, Yemen Red Sea — monitored simultaneously on one platform.

📊

Confidence Bands

Every probability shown with its uncertainty range. Decision-makers see how firm the assessment is, not just what it says.

🔍

Full Audit Trail

Every probability shift logged. What changed. Which signal. How much it moved the needle. Nothing unexplained.

🗺️

Map Intelligence

Instability, Waterways, Economic, Cyber, Natural and Military overlays — geospatial context for every assessment.

🤖

AI Insights Brief

Auto-generated World Brief and Strategic Posture — executive-ready summary from live data.

📡

Live Data Feeds

ACLED, GDELT, USGS and custom OSINT — continuously ingested, weighted and surfaced to the decision layer.

// REQUEST A DEMONSTRATION

See DRISHTI tracking a live theatre relevant to your environment. Our team will walk you through the full escalation model and audit trail.

IC&C

AI-Powered Intrusion Detection & Command and Control Platform

The Security Reality

Perimeters generate alarms.
None of them produce decisions.

Modern perimeter security generates enormous volumes of alerts — from OFC sensors, electric fences, cameras and motion detectors. Operators are flooded with alarms they must manually verify, correlate and act upon. The result is alert fatigue, delayed response, and security gaps that adversaries exploit.

"The challenge is not detecting the intrusion. It is knowing — with confidence — that it is real, where exactly it is, and what to do about it in the next 90 seconds."

— IC&C Design Philosophy

🚨

Alert Flood

OFC, fence, camera and motion sensors all alarm independently — no system correlates them into a single, confident threat picture.

📍

No Localisation

Operators know something happened — not where exactly, not from which sensors, and not with what confidence level.

🔀

Fragmented Response

Camera operators, security teams, drone pilots and command staff work in separate systems with no unified workflow.

82%

Fusion confidence — when OFC, fence and camera corroborate simultaneously

<90s

Target from alarm to drone dispatch with SOP-driven operator workflow

2-tier

Field Control Platform + Master Control Platform — site to command

GPS

Precise geo-coordinates for every intrusion event, displayed live on operator map

The Gap

Traditional security systems detect.
They do not decide.

The gap in perimeter security is not sensor coverage — it is the intelligence layer between detection and response. Without multi-sensor fusion, GPS localisation, automated camera cueing, and SOP-driven workflows, operators are left to improvise under pressure.

No multi-sensor correlation. OFC, electric fence and cameras alarm in separate systems. No platform fuses them into a single confidence-scored incident.

No automatic camera cueing. When an alarm fires, the nearest relevant camera does not activate automatically — a human must find and switch to it manually, losing precious seconds.

No GPS-referenced response. Alarms lack precise coordinates — security teams must physically search a zone rather than respond to an exact location.

No SOP-driven operator workflow. Response depends entirely on individual training and memory — inconsistent, unauditable and unrepeatable.

No unified command view. Field operators, supervisors and central command work in disconnected silos — with no common operational picture across sites.

The Platform

IC&C closes the response gap.
Two platforms. One unified command.

IC&C delivers a two-tier command architecture — a Field Control Platform at site level and a Master Control Platform at central command. Together they transform sensor alarms into confident, GPS-referenced, SOP-driven responses — across any perimeter, at any scale.

🎯

Field Control Platform (FCP)

Site / Sector Level · Operator Console

Real-time intrusion detection with precise GPS localisation

Multi-sensor alarm fusion — OFC + fence + camera correlated

Automatic camera cueing to nearest relevant feed

SOP-driven operator workflow — Acknowledge → Verify → Dispatch

Drone coordination interface with mission progress tracking

Full incident logging and audit trail

↕

SECURE API

🌐

Master Control Platform (MCP)

Central Command · Supervisor Console

Multi-site GIS overview — all connected sites on one map

Central alarm aggregation across all Field Control Platforms

SOP oversight — visibility into field workflow execution

Cross-site escalation and supervisory intervention

Alarm trend analytics and response-time KPIs

Admin governance — roles, access and audit management

SENSORS IN SCOPE

OFC Sensors

X/Y position · Latitude / longitude · Intrusion type classification

Electric Fence

Zone-wise alert · Power status · Breach location coordinates

Camera Network

Live video feed · Auto-cued on alarm · PTZ and thermal support

AI-ENABLED PLATFORM

AI-Assisted Intrusion Detection

Multi-sensor correlation algorithms that improve detection accuracy and reduce false alarms.

Predictive Security Analytics

ML models identifying intrusion patterns, vulnerable segments and high-risk time windows.

AI Command Interface

Suggests response actions, summarises incidents and assists command operators during emergencies.

The Dashboard

From standby to response.
Every second visible. Every action auditable.

IC&C — PERIMETER COMMAND & CONTROL · LIVE SATELLITE VIEW · 300m PERIMETER · 4 TOWERS · 8 CAMERAS · 1 DRONE

🔍

Multi-Sensor Fusion

OFC + electric fence + camera correlated into a single confidence-scored incident. No single sensor triggers alone.

📍

GPS Localisation

Every intrusion event carries precise coordinates — displayed instantly on the GIS map for immediate response targeting.

📷

Auto Camera Cueing

The nearest relevant camera activates automatically on alarm — no manual switching, no lost seconds.

📋

SOP Workflows

Acknowledge → Verify → Dispatch — every operator follows the same auditable sequence, every time.

🛸

Drone Coordination

Eagle-1 dispatched to INC-001 coordinates. Mission progress, battery, position and ETA tracked in real time.

🏛

Central Command

MCP gives supervisors a live multi-site view — drill down to any incident, monitor SOP execution, escalate cross-site.

// REQUEST A DEMONSTRATION

See IC&C run a live intrusion simulation — from sensor alarm to drone dispatch — on a perimeter relevant to your security environment.

DSS-SOS

Decision Support & Safety Operating System for Industries

The Operational Reality

In emergencies, every system talks.
None of them explains.

Oil and gas facilities are among the most heavily instrumented environments in the world. 10,000+ sensors per facility. DCS, SIS, Fire & Gas systems that detect and alarm with precision. And yet — when something goes wrong, the operator is left alone to make the most consequential decisions of their shift, with incomplete information, under extreme time pressure.

"The challenge is not DETECTION. It is DECISION MAKING."

— DSS-SOS Design Philosophy

GAS LEAK SCENARIO — THE NARROW DECISION WINDOW

                SAFE
                EARLY WARNING
                HIGH RISK
              

▲ 18% LEL
Current Reading

                0%10%40%100%
              

This is the narrow window where the right decision prevents escalation.

T + 0

Gas detected — no decision made

T + 1 min

Concentration crosses 20% LEL

T + 3 min

Ignition probability increases sharply

T + 8 min

Potential unit-level escalation

🚨

Alarm Floods

10–30 alarms may trigger from a single abnormal event. No prioritisation. No ranking. No guidance.

🔧

Equipment-Centric

Systems are built around equipment — not the operator. Every system talks about its own domain. None synthesises across them.

⏱️

Experience Under Pressure

Response quality depends on individual experience. Under stress, even trained operators hesitate. Minutes determine outcomes.

The Missing Layer

You have detection.
You are missing the decision layer.

Distributed systems control the plant. But no system truly guides the operator. You already have everything needed for detection — what is missing is the intelligence layer that converts all of that into a ranked, time-critical decision.

WHAT YOU HAVE

Sensors, DCS, SIS, Fire & Gas systems

SOPs and emergency procedures

Trained operations and response teams

10,000+ data points per facility

WHAT IS MISSING

A system that answers three questions:

What should be done first?

What happens if action is delayed?

Which action best limits escalation?

WHERE DSS-SOS SITS

DETECTION & CONTROL

10,000+ sensors · DCS · SIS · Fire & Gas

▼

DSS-SOS — DECISION INTELLIGENCE LAYER

Situation synthesis · Action prioritisation · Escalation timelines · Confidence indicators

THE GAP FILLED

▼

ACTION & RESPONSE

SOPs · Emergency response teams · Operator decisions

The Solution

From uncertainty to meaningful decisions.
Every second saved reduces escalation risk.

DSS-SOS is a decision intelligence layer that sits above your existing systems. It integrates real-time operational data, simplified hazard estimation, SOP logic and AI-based prioritisation — and produces ranked, actionable guidance. Not just alerts.

HOW DSS-SOS THINKS — FROM DATA TO DECISIONS

📡

Sensors & Systems

DCSSISFire & GasOPC-UAModbusONVIFMQTTREST APIs

▼

🔍

Context & State

What is abnormal?Where is it happening?How fast is it changing?

▼

🧠

Decision Space

What actions are possible?What happens if we delay?What escalates first?

▼

✅

Prioritised Actions

What to do firstWhat can waitWhat prevents escalation

PLATFORM COMPONENTS

UI/UX

Operator-centric interfaces designed for stress conditions. Clear. No jargon. Only operational guidance.

Visualisation

Zone-based situational awareness and escalation view. Visual overlays of affected areas.

Sensor Integration

Real-time integration with sensors, DCS, SIS, Fire & Gas. Vendor-agnostic, non-intrusive.

Simulation

Hazard estimation and escalation timelines. Fire dynamics modelling. What-if analysis.

Training

Scenario replay, shift drills, SOP validation. Pre-job briefs. Offline rehearsal capability.

Secure Language Model

On-premises SLM chatbot. Query your procedures and operational knowledge offline.

ROLE OF AI IN DSS-SOS

Rules alone are not enough. True incidents are multi-domain: process + physical + visual + procedural. AI handles what rules cannot.

Correlate multiple sensor streams

Detect abnormal patterns early

Rank actions by urgency & impact

Probabilistic scoring of actions

Experience-based weighting of decisions

Continuous refinement using ops data

"AI supports human decisions — it does not replace them."

The Outcome

What operators see.
No jargon. Only clear operational guidance.

DSS-SOS OPERATOR CONSOLE

● FIRE ALERT

🔥 GAS LEAK — COMPRESSOR UNIT 4 · TRAIN B

14:07:22 · SOP: GL-04 · Confidence: HIGH

RANKED ACTIONS

Emergency Shutdown — Train B

▸ Close SDV-401 · Isolate fuel gas

Activate Deluge System — Zone 4

▸ Confirm activation valve DV-04

Evacuate Personnel — 50m radius

▸ Muster at Assembly Point B

Isolate Adjacent Process Units

▸ Close XV-402, XV-403

Notify Control Room & ERC

▸ Escalate to Area Authority

ESCALATION TIMELINE

● LIVE

COMPRESSOR UNIT 4 — TRAIN B · REFINERY COMPLEX

T+0

Gas concentration at 18% LEL — seal failure suspected

T+1m

Crosses 20% LEL — ESD initiation window

T+3m

Vapour cloud forming — ignition risk elevated

T+5m

Thermal exposure risk to crude distillation unit

T+8m

Plant-wide escalation risk — full evacuation

CONFIDENCE INDICATOR

HIGH · Gas detector + pressure sensor + CCTV correlated · ESD recommended

🎯

Ranked Actions

Every alarm produces a prioritised list. P1 first. Always. No operator left guessing what matters most.

⏱

Escalation Timelines

The system shows what happens at T+1, T+3, T+8 — so decisions are made before the window closes.

📊

Confidence Indicators

Every recommendation carries a confidence score. Operators know how certain the system is — and why.

📋

SOP-Driven Workflow

Procedures are embedded — not in a manual on a shelf. Operators follow site-specific SOPs in real time.

🧠

Built With Operators

"The best decision-support systems are built with operators, not for them." We embed tacit knowledge, not just rules.

🔄

Continuously Improving

AI learns from every incident and drill. The system improves its prioritisation the longer it runs at your facility.

// REQUEST A DEMONSTRATION

See DSS-SOS run a live scenario — loss of containment, fire response, or a scenario specific to your facility. From data to decisions, when it matters most.

DT&AS

Distributed Temperature & Acoustic Sensing — Persistent Infrastructure Surveillance

The Challenge

Critical infrastructure spans
hundreds of kilometres.
Point sensors see almost none of it.

Pipelines, perimeter fences, power cables, rail corridors and subsea assets stretch across vast distances. Traditional sensor systems — placed at discrete points — leave enormous blind spots between them. An intrusion, leak or temperature anomaly anywhere in those gaps goes undetected until it becomes a crisis.

"The fibre-optic cable you already have buried alongside your pipeline is also a sensor. A 50km sensor, measuring everything happening along its entire length, continuously."

— DT&AS Design Philosophy

📍

Point Sensor Gaps

Conventional sensors monitor specific locations. Everything between them is invisible — and that is where threats and failures occur.

⏳

Delayed Detection

By the time a leak, intrusion or hotspot reaches a point sensor, significant damage has already occurred. DT&AS detects at inception.

💰

Infrastructure Cost

Deploying conventional sensors across 50km of pipeline or perimeter is prohibitively expensive. DT&AS uses the fibre already in the ground.

50km+

Continuous sensing along a single fibre-optic cable

Spatial resolution — events located to within one metre

Technologies in one — acoustic and temperature, same cable

24/7

Continuous passive monitoring — no moving parts, no maintenance windows

Distributed Acoustic Sensing

Every vibration. Every footstep.
Detected. Located. Classified.

Distributed Acoustic Sensing turns a standard fibre-optic cable into a continuous microphone array. Laser pulses sent along the fibre detect the faintest changes in backscattered light caused by acoustic vibrations — from footsteps on a perimeter fence to a pump cavitating inside a pipeline to a vehicle approaching a critical corridor.

HOW DAS WORKS

Light Source

Pulsed laser injected into fibreNanosecond pulses

▼

Rayleigh Backscattering

Acoustic vibrations disturb backscattered lightShift detected at interrogator

▼

Signal Processing

Pattern recognitionEvent classificationLocation calculation

▼

Alert Output

Event typeGPS coordinatesConfidence scoreRecommended action

WHAT DAS DETECTS

🚶

Perimeter Intrusion

Footsteps, fence cutting, vehicle approach — detected and classified up to 50km

🔧

Third-Party Interference

Digging, drilling or mechanical activity near buried pipelines — detected before damage

💧

Pipeline Leak Acoustics

Leak-induced acoustic signatures detected and located to within metres along the pipe

⚙️

Equipment Anomalies

Pump cavitation, valve chatter, bearing wear — acoustic fingerprints detected early

🚗

Traffic & Seismic

Vehicle classification, rail monitoring, ground movement along critical corridors

DAS — LIVE ACOUSTIC FEED · 47.3km MONITORED

● ACTIVE

14:22:07DAS HIGHAcoustic burst · KM 12.3 · Pattern: probable excavation · Alert dispatched

14:21:44DAS MEDSustained vibration · KM 28.7 · Pattern: vehicle · Monitoring

14:20:58DAS INFOLow-frequency signal · KM 6.1 · Pattern: pump cavitation signature

Distributed Temperature Sensing

Every degree. Every metre.
Continuous thermal profile along the entire asset.

Distributed Temperature Sensing uses Raman scattering to measure temperature at every point along the fibre — not just at sensor locations. The result is a continuous, real-time thermal map of your asset. A hotspot developing at kilometre 31.4 of a pipeline is as visible as one right next to the control room.

HOW DTS WORKS

Light Source

Pulsed laser along fibreContinuous interrogation

▼

Raman Scattering

Anti-Stokes backscatter is temperature-sensitiveRatio analysis gives absolute temperature

▼

Thermal Profiling

Continuous temperature mapResolution to ±0.1°CSpatial resolution 1m

▼

Alert Output

Hotspot locationRate of changeThreshold breach alert

WHAT DTS DETECTS

🔥

Fire & Hot Spot Detection

Early-stage thermal anomalies detected before visible flame — in tunnels, cable trays, conveyor belts

🛢️

Pipeline Leak Detection

Hydrocarbon leaks cause localised temperature drop — detected and located to within metres

⚡

Cable & Electrical Monitoring

Overheating in power cables, transformer faults, conductor anomalies — before failure

🏭

Process Temperature Profiling

Continuous temperature monitoring along reactors, vessels and heat exchangers

🌱

Environmental Monitoring

Soil temperature, permafrost changes, subsurface thermal mapping along buried assets

DTS — LIVE THERMAL PROFILE · PIPELINE KM 0 → KM 47

HOTSPOT

KM 25.8 · 94°C · Rate: +2.1°C/min

BASELINE

42–46°C across 45.1km of asset

RESOLUTION

±0.1°C · 1m spatial · Updated every 30s

Where We Deploy

One fibre. Two technologies.
Deployed across your most critical assets.

DT&AS can be deployed wherever fibre-optic cable runs — or can be installed alongside existing infrastructure. The same interrogator unit drives both DAS and DTS, providing acoustic and thermal intelligence from a single deployment.

🛢️

Oil & Gas Pipelines

Leak detection, third-party interference, cathodic protection monitoring and flow assurance across onshore and offshore pipelines.

DASDTS

🏭

Refineries & Industrial

Continuous temperature profiling of process equipment, fire detection in cable trays and tunnels, equipment anomaly detection.

DTS

🛡️

Border & Perimeter Security

Continuous acoustic monitoring of fenced perimeters, border corridors and restricted zones — intrusion detected and located in real time.

DAS

⚡

Power Cables & Grid

Thermal monitoring of high-voltage cables, early detection of insulation degradation and overheating before fault or outage.

DTS

🚂

Railways & Transport

Track monitoring, intrusion detection, landslide and rockfall early warning, axle bearing temperature along rail corridors.

DASDTS

🌊

Subsea & Offshore

Subsea pipeline monitoring, riser thermal profiling, anchor drag detection and seabed acoustic surveillance.

DASDTS

// REQUEST A DEMONSTRATION

See DT&AS deployed on a scenario relevant to your infrastructure — pipeline, perimeter, power cable or process facility. Our team will walk you through a live sensor feed and alert generation.

Digital Twin

Digital Twin & Simulation

Capability Overview

Your platform. Replicated.
Ready to stress-test before the mission.

Medjay builds high-fidelity digital replicas of physical systems — ships, facilities, command environments — so operators can train, rehearse and validate decisions without real-world risk. When live sensor data feeds in, the twin becomes a continuous decision-support tool, not just a model.

▸Real-time digital replication of physical systems, assets and operational environments.

▸Scenario-based simulation for training, testing, mission rehearsal and operational validation.

▸Integration with live sensor data for deeper operational visibility and informed decision support.

▸Predictive modelling to assess system behaviour under stress, disruption and failure conditions.

The gap between a plan and its execution is where decisions fail. Digital twin closes that gap.

🧭

Model

Create a structured digital representation of assets, systems, spaces and workflows.

🎮

Simulate

Run operational scenarios, stress cases and response sequences in a safe environment.

📈

Improve

Use simulation outcomes to refine doctrine, training, design and operational readiness.

🚢

Defence & Maritime

Mission rehearsal, damage-control scenarios, system familiarisation and decision support for complex operational environments.

🏭

Industrial Systems

Model plants, utilities and process environments for training, disruption simulation and response planning.

🛡️

Security Operations

Simulate intrusion paths, sensor coverage, response timelines and command workflows for preparedness and optimisation.

🧠

Decision Support

Integrate live or historical data into simulation environments to support training, testing and what-if analysis.

👥

Training & Rehearsal

Enable operators and teams to rehearse procedures, validate response logic and build confidence before real-world deployment.

⚙️

Design Validation

Assess layouts, procedures and system interactions digitally before implementation, modification or scale-up.

// REQUEST A DEMONSTRATION

See how Medjay can configure a digital twin and simulation environment aligned to your platform, facility or mission requirement.

From Sensing
to Decision Superiority

Intelligence Systems
for Critical Environments

Decision Intelligence Platforms

Fire. Flooding. Simultaneously.
Four minutes to make the right call.

Traditional damage control
was not built for this complexity.

FFDCS closes the decision gap.
Superior situational awareness. AI-enabled command.

Faster decisions. Protected crew.
Ships that Float, Move and Fight.

Every system sees a piece.
None of them see the picture.

Current systems generate alerts.
None recommend actions.

PRAVA closes the decision gap.
One platform. Four domains. Offline-first.

One core. Four mission packages.

Edge-first. Modular. Air-gapped.
Built for environments where cloud is not an option.

Decision-makers are drowning
in signals they cannot act on.

Intelligence products tell you what happened.
Not what is next.

DRISHTI answers three questions.
Every briefing. Every theatre. Live.

Three explicit layers.
Signals. Inference. Decision.

Built for decision-makers.
Not for analysts writing reports.

Perimeters generate alarms.
None of them produce decisions.

Traditional security systems detect.
They do not decide.

IC&C closes the response gap.
Two platforms. One unified command.

From standby to response.
Every second visible. Every action auditable.

In emergencies, every system talks.
None of them explains.

You have detection.
You are missing the decision layer.

From uncertainty to meaningful decisions.
Every second saved reduces escalation risk.

What operators see.
No jargon. Only clear operational guidance.

Critical infrastructure spans
hundreds of kilometres.
Point sensors see almost none of it.

Every vibration. Every footstep.
Detected. Located. Classified.

Every degree. Every metre.
Continuous thermal profile along the entire asset.

One fibre. Two technologies.
Deployed across your most critical assets.

Your platform. Replicated.
Ready to stress-test before the mission.

Where We Operate

Why Medjay

Ready to Deploy?

Our Core
Philosophy.

From Sensingto Decision Superiority

Intelligence Systemsfor Critical Environments

Decision Intelligence Platforms

Fire. Flooding. Simultaneously.Four minutes to make the right call.

Traditional damage controlwas not built for this complexity.

FFDCS closes the decision gap.Superior situational awareness. AI-enabled command.

Faster decisions. Protected crew.Ships that Float, Move and Fight.

Every system sees a piece.None of them see the picture.

Current systems generate alerts.None recommend actions.

PRAVA closes the decision gap.One platform. Four domains. Offline-first.

One core. Four mission packages.

Edge-first. Modular. Air-gapped.Built for environments where cloud is not an option.

Decision-makers are drowningin signals they cannot act on.

Intelligence products tell you what happened.Not what is next.

DRISHTI answers three questions.Every briefing. Every theatre. Live.

Three explicit layers.Signals. Inference. Decision.

Built for decision-makers.Not for analysts writing reports.

Perimeters generate alarms.None of them produce decisions.

Traditional security systems detect.They do not decide.

IC&C closes the response gap.Two platforms. One unified command.

From standby to response.Every second visible. Every action auditable.

In emergencies, every system talks.None of them explains.

You have detection.You are missing the decision layer.

From uncertainty to meaningful decisions.Every second saved reduces escalation risk.

What operators see.No jargon. Only clear operational guidance.

Critical infrastructure spanshundreds of kilometres.Point sensors see almost none of it.

Every vibration. Every footstep.Detected. Located. Classified.

Every degree. Every metre.Continuous thermal profile along the entire asset.

One fibre. Two technologies.Deployed across your most critical assets.

Your platform. Replicated.Ready to stress-test before the mission.

Where We Operate

Why Medjay

Ready to Deploy?

Our CorePhilosophy.

From Sensing
to Decision Superiority

Intelligence Systems
for Critical Environments

Fire. Flooding. Simultaneously.
Four minutes to make the right call.

Traditional damage control
was not built for this complexity.

FFDCS closes the decision gap.
Superior situational awareness. AI-enabled command.

Faster decisions. Protected crew.
Ships that Float, Move and Fight.

Every system sees a piece.
None of them see the picture.

Current systems generate alerts.
None recommend actions.

PRAVA closes the decision gap.
One platform. Four domains. Offline-first.

Edge-first. Modular. Air-gapped.
Built for environments where cloud is not an option.

Decision-makers are drowning
in signals they cannot act on.

Intelligence products tell you what happened.
Not what is next.

DRISHTI answers three questions.
Every briefing. Every theatre. Live.

Three explicit layers.
Signals. Inference. Decision.

Built for decision-makers.
Not for analysts writing reports.

Perimeters generate alarms.
None of them produce decisions.

Traditional security systems detect.
They do not decide.

IC&C closes the response gap.
Two platforms. One unified command.

From standby to response.
Every second visible. Every action auditable.

In emergencies, every system talks.
None of them explains.

You have detection.
You are missing the decision layer.

From uncertainty to meaningful decisions.
Every second saved reduces escalation risk.

What operators see.
No jargon. Only clear operational guidance.

Critical infrastructure spans
hundreds of kilometres.
Point sensors see almost none of it.

Every vibration. Every footstep.
Detected. Located. Classified.

Every degree. Every metre.
Continuous thermal profile along the entire asset.

One fibre. Two technologies.
Deployed across your most critical assets.

Your platform. Replicated.
Ready to stress-test before the mission.

Our Core
Philosophy.