TIDS · TRAIN INTELLIGENT DETECTION SYSTEM

Intelligent Train Obstacle Detection Solution

Give Trains Independent Forward-environment Perception

LiDAR, industrial cameras, optional millimeter-wave radar and intelligent algorithms actively perceive trains, people, intruding objects and signal states during normal operation, degraded modes, line inspection, commissioning and depot shunting.

Explore the Complete Solution Consult a Solution Specialist

01 Autonomous Positioning02 Obstacle Detection03 Signal Recognition04 Risk Output

CUSTOMER VALUE

Customer Value Summary

Give train crews continuous forward risk alerts to support anomaly detection and response.

Support risk detection

Detect people, vehicles and obstacles ahead to supplement the driver's lookout and identify potential hazards.

Support varied operating scenarios

Configure alerts for service operation, degraded operation, testing and shunting to support crews across operating conditions.

Support event review

Link targets, positions and alert records so operators can review events and inform operational improvements.

INDEPENDENT ONBOARD PERCEPTION

Add a Continuously Forward-looking Pair of Eyes beyond Existing Systems

Centered on onboard perception, TIDS does not rely on a single vehicle sensor or signaling position. It combines active and passive sensing to identify objects along the train's potential path and assess risk.

Complete DeliveryComplete TIDS Intelligent Train Obstacle Detection System

OPERATIONAL GAPS

Four Protection Gaps in Complex Rail Operations

Intruding Objects Are Hard to Detect in Time

When drill rods, blast doors, trees, cables or other objects enter the safety clearance, conventional systems lack direct perception of non-vehicle obstacles.

Degraded Operation Relies on Human Lookout

When signaling or vehicle systems degrade, safety depends more heavily on drivers and dispatchers, affecting response time and consistency.

Incomplete Protection during Commissioning and Shunting

During line inspection, trial runs, cross-line operation and depot shunting, existing systems may not fully protect against trains ahead, buffer stops and personnel.

Complex Environments Challenge Perception Stability

Tunnel lighting transitions, curves, gradients, turnouts, rain, snow and fog continually change visibility and target appearance.

OPERATING SCENARIOS

Cover Normal Operation, Degraded Modes, Commissioning and Shunting

Configure target recognition, alerts and vehicle-interface strategies according to operating rules and the protection boundaries of existing systems.

Normal Operation

Continuously perceive intruding objects, people in platform track areas, and buffer stops on sidings or terminal tracks.

Degraded Operation

Provide independent perception of trains ahead, signals, buffer stops and track objects when signaling or vehicle systems are degraded.

Line Inspection & Commissioning

Support train-ahead protection and detection of people and small objects during commissioning, reducing the burden of manual lookout.

Depot Shunting

Cover depot doors, buffer stops, vehicles ahead, trackside personnel and abnormal obstacles along shunting paths.

INTEGRATED ARCHITECTURE

Coordination across Perception, Computing and Vehicle Interfaces

Multiple sensors collect environmental data on a shared time base. The main computer performs fused perception and risk assessment, then outputs results to the driver or vehicle system.

PERCEPTION

Industrial CamerasLiDARMillimeter-wave Radar Optional

↓

EDGE INTELLIGENCE

Synchronization & FusionAutonomous PositioningTarget RecognitionRisk Calculation

TIDS Main Computer

↓

VEHICLE INTERACTION

DMI DisplayAudible & Visual AlertsVehicle BusBraking Interface

TIDS autonomous train perception technologies and system components — Documented Key Technologies and Architecture for Autonomous Train Perception

CORE CAPABILITIES

From Seeing a Target to Producing Actionable Information

Perception-based Autonomous Positioning

Match point clouds to a high-definition map for real-time autonomous train positioning and a spatial reference for filtering targets along the route.

Forward Obstacle Detection

Identify trains, people, buffer stops and foreign objects entering the safety clearance along the train's potential path.

Signal Aspect Recognition

Recognize red, yellow, green and other signal aspects during mainline operation or shunting to support driving and protection logic.

Risk Calculation & Interface Output

Calculate risk from vehicle speed, braking performance and obstacle distance, then output warnings and intervention information through the DMI, vehicle bus or braking interface.

<1sResponse to Sudden Events

≥300mTrain-ahead Warning Distance

≥120mPersonnel Detection Distance

≥50m30×30 cm Object Detection Distance

These are representative values under the documented test conditions; distance values correspond to straight track. Actual performance depends on vehicle type, line, weather, installation and system configuration. Project specifications and acceptance results prevail.

REAL-TIME DETECTION

Detect Targets and Determine Whether They Affect the Train's Path

Using train position, track geometry and safety clearance, the system filters targets relevant to the train from complex point clouds and images, reducing interference from adjacent-line vehicles and objects outside the clearance.

Dynamic tracking and distance calculation for trains ahead
Detection of people, buffer stops and small objects
Path assessment on straight track, curves, gradients and turnouts
Continuous perception by day, night, rain and snow

TIDS forward train recognition interface

FLEXIBLE OPERATING MODES

Map-based and Map-free Operating Modes

Map-based Mode

Use a 3D high-definition map for autonomous positioning, obstacle detection and signal recognition on mapped lines.

Line position reference · Path-associated detection

Map-free Mode

Provide forward obstacle detection before an HD map is available, supporting rapid deployment for commissioning, temporary operation and cross-line running.

Reduced mapping prerequisites · Temporary scenario configuration

STANDARD WORKFLOW

Six Steps Form a Complete Perception Loop

Synchronized Multi-source Capture

Cameras, LiDAR and optional millimeter-wave radar capture the forward environment synchronously.

Autonomous Positioning

Determine the train's real-time position, direction and target trajectories on the line.

Target Recognition

Distinguish trains, people, buffer stops, signals and other foreign objects.

Clearance Assessment

Determine whether a target enters the protected area based on the line and travel path.

Risk Calculation

Combine speed, distance and vehicle performance to define warning or braking-intervention conditions.

Output & Recording

Output results to the driver or vehicle interface and retain operational data for review.

SYSTEM DELIVERY

Configure the Complete System to Vehicle Type and Project Scope

Equipment quantity, installation positions, interfaces and braking coordination must be determined from vehicle space, line conditions and safety requirements.

Onboard Sensor Unit

Industrial cameras, LiDAR and optional millimeter-wave radar configured to the vehicle and available installation space.

View Product Details ↗

TIDS Main Computer

Performs sensor synchronization, fused perception, autonomous positioning, target recognition and risk calculation.

View Product Details ↗

DMI Driver Display

Presents target type, distance, speed, line status and graded warnings to the driver.

View Product Details ↗

Vehicle & Signaling Interfaces

MVB/TRDP bus, signaling network, audible/visual alerts and braking interfaces can be configured to vehicle conditions.

TIDS detection in tunnels, curves, turnouts and elevated sections

ENVIRONMENT ADAPTABILITY

Designed for Real Lines, Not a Single Laboratory Environment

Documentation shows continued data accumulation in tunnels, viaducts, surface sections, curves, gradients and turnouts, with validation under lighting transitions, backlight, rain and snow.

Track Geometry: Straight / Curve / Gradient / Turnout / Viaduct
Lighting Conditions: Unlit / Light transitions / Front light / Backlight / Night
Weather Conditions: Rain / Snow / Fog and other project-defined conditions

FIELD FINDINGS

Make Potential Risks Visible before a Collision

TIDS detects tree intrusion during commissioning — **Tree Intrusion**Detect trees intruding into the operating clearance during commissioning

TIDS detects a train ahead inside a tunnel — **Train in Tunnel**Detect a vehicle target ahead in an operating environment

TIDS detects a buffer stop and issues an overspeed alert — **Buffer-stop Protection**Recognize the buffer stop and issue a risk prompt during depot return

APPLICATION EXPERIENCE

Multi-line Experience and Engineering Practice

The following are representative line applications from the available materials. Actual project scope is subject to authorized releases and contract terms.