Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC TS 62580-2: Railway Applications — Onboard Multimedia — Part 2: Video Surveillance
System architecture, functional requirements, and performance specifications for video surveillance systems on railway rolling stock
IEC TS 62580-2, published as a Technical Specification by IEC Technical Committee 9 (Electrical Equipment and Systems for Railways), defines the system architecture, functional requirements, and performance specifications for video surveillance systems installed on railway rolling stock. As railway operators worldwide face increasing demands for passenger safety, security incident investigation, and operational monitoring, onboard video surveillance has become an essential component of modern train design. This standard provides a unified framework ensuring that CCTV systems across different train types — from high-speed trains and metro vehicles to light rail and regional trains — meet consistent performance levels for image quality, recording reliability, data security, and integration with railway communication networks.
IEC TS 62580-2 addresses the complete video surveillance chain on railway vehicles: from camera selection and placement, through video encoding and onboard recording, to data offloading and integration with ground-based security management systems. The standard covers both visible-light and thermal imaging cameras, stationary and Pan-Tilt-Zoom (PTZ) camera types, and supports integration with train management systems, public address systems, and emergency communication interfaces.
System Architecture and Camera Specifications
The standard defines a distributed system architecture for onboard video surveillance. The system comprises four functional tiers: the acquisition tier (cameras and image sensors), the processing tier (encoding, analytics, and local storage controllers), the storage tier (onboard digital video recorders with redundant storage), and the communication tier (offloading interfaces and real-time transmission links). Cameras are connected to local processing units via Ethernet (100BASE-TX minimum, with Power over Ethernet for simplified cabling) or coaxial cable for legacy installations. The train backbone network, based on IEC 61375 (Train Communication Network), interconnects the processing units and provides the communication path to the driver display, train control system, and wireless offloading gateway.
Camera specifications are among the most detailed parts of the standard. Minimum requirements include: resolution of at least 1280 x 720 pixels (720p) for general surveillance and 1920 x 1080 (1080p) for critical areas such as door zones and driver cabs; minimum frame rate of 6 fps for general areas and 25 fps for areas requiring detailed motion analysis; minimum sensitivity of 0.05 lux for colour operation and 0.01 lux for monochrome operation; automatic day/night switching with IR-cut filter; wide dynamic range of at least 100 dB for environments with high contrast lighting such as tunnel exits and station platforms; and minimum Mean Time Between Failures of 50,000 hours for the camera module. The standard also mandates integrated heater and blower for cameras mounted externally to prevent condensation and ice formation in cold climates.
| Application Zone | Min. Resolution | Min. Frame Rate | Coverage | Special Requirements |
|---|---|---|---|---|
| Driver cab | 1080p | 25 fps | Full cab area, operator seat | Audio recording, IR illumination |
| Passenger saloon | 720p | 6 fps (general), 25 fps (incident) | 95% of floor area, all entrances | Wide dynamic range >= 100 dB |
| Door zones | 1080p | 25 fps | Full door opening, platform interface | Thermal camera optional, IR illumination |
| Gangway connections | 720p | 12 fps | Full connection area | Anti-vibration mounting |
| External (platform side) | 1080p | 25 fps | Platform-train interface, door area | Weatherproof (IP66+), heater, wiper |
| Technical areas | 720p | 6 fps | Equipment cabinets, power distribution | Low-light performance, 0.01 lux |
One of the most challenging engineering aspects of railway CCTV is the harsh operating environment: cameras must function reliably across temperature ranges from -40 deg C (arctic operations) to +70 deg C (enclosed equipment cabinets in summer), under continuous vibration up to 5.7 m/s2 RMS per IEC 61373 Category 1 Class B, and in electromagnetic environments up to 100 V/m field strength as specified in IEC 62236-3-2. Connectors must be rated for at least 500 mating cycles and sealed to IP67 to prevent moisture ingress during pressure washing of the train exterior.
Video Recording, Storage, and Data Retrieval
The standard defines a pre-event and post-event recording model. The onboard digital video recorder continuously captures video in a circular buffer, retaining at minimum 30 seconds of pre-event video and at least 10 minutes of post-event video at full frame rate. Recording is triggered by configurable events: train door opening/closing, emergency brake activation, passenger emergency communication activation, train speed above threshold, time-scheduled recording, and manual activation by the driver or train crew. The recording trigger events and associated video segments are indexed in a searchable database with metadata including timestamp (synchronized to GPS or IRIG time), train ID, car number, camera ID, event type, and train speed at the time of recording.
Video compression uses H.264 (AVC) High Profile as the mandatory baseline codec, with H.265 (HEVC) recommended for new installations to reduce storage requirements by approximately 40-50% at equivalent quality. The standard specifies constant bitrate encoding with a maximum bitrate of 8 Mbps per camera for 1080p at 25 fps, and 4 Mbps per camera for 720p at 6 fps. Storage capacity must accommodate at least 72 hours of continuous recording from all cameras in the standard operating configuration. For a typical metro train with 16 cameras, this requires a minimum onboard storage capacity of approximately 500 GB using H.264 encoding at the specified bitrates, or approximately 300 GB using H.265 encoding. Redundant storage in RAID 1 (mirrored) configuration is mandatory, with automatic failover and recording continuity verification through continuous write-verify cycles. The recording system must be protected against data loss during power loss events through a supervised graceful shutdown procedure, with an uninterruptible power supply providing at least 30 seconds of operation after main power failure to complete pending write operations and close all video files cleanly.
Modern railway CCTV systems implementing IEC TS 62580-2 can store over 30 days of continuous recording using compression-optimized H.265 encoding and adaptive bitrate management that reduces recording quality during low-activity periods (night-time, depot stabling) while maintaining full quality during high-traffic periods. Combined with edge-based video analytics for motion-triggered recording optimisation, storage requirements can be reduced by an additional 50-70% compared to continuous recording at constant quality, while still ensuring that all security-relevant events are captured at the highest quality level.
Engineering Design Insights for Railway CCTV Systems
When designing video surveillance systems in accordance with IEC TS 62580-2, engineers must address several key challenges unique to the railway environment. First, video data offloading from the train to the ground-based security management system requires careful planning of the communication infrastructure. The standard defines three offloading methods: wireless offloading via Wi-Fi (IEEE 802.11n/ac, at depot arrival), cellular offloading (4G LTE or 5G NR for real-time access to selected camera streams), and physical offloading via removable storage media or wired Ethernet connection during maintenance. For wireless offloading at depots, the standard recommends that the system be capable of transferring the complete recording from a single train within 30 minutes of arrival, requiring a sustained data rate of at least 50 Mbps for a typical 16-camera configuration with 72 hours of storage. Multiple trains arriving simultaneously at the depot must be supported through sectorized Wi-Fi access points with automatic load balancing.
Second, the integration of video surveillance with other train systems requires standardized interfaces. The standard mandates a data interface between the CCTV system and the Train Control and Management System (TCMS) for event-driven recording triggers and system health monitoring. The CCTV system must transmit a periodic heartbeat signal to the TCMS indicating operational status, storage utilisation, and camera health. The standard also recommends integration with the passenger emergency communication system: when a passenger activates an emergency intercom, the nearest camera view is automatically displayed on the driver’s surveillance monitor and the relevant video segment is permanently retained to support incident investigation. For interoperability across different manufacturers, the standard references ONVIF Profile S for camera communication and RTSP/RTP for video streaming, with H.264 video encapsulated in MP4 container format for recorded segments.
Third, data security and privacy compliance are critical requirements. Recorded video data containing passenger images is subject to data protection regulations including the EU General Data Protection Regulation. The standard requires that video data be encrypted at rest using AES-256 encryption, with access controlled through role-based authentication (driver, investigator, maintenance, security administrator roles with graduated access permissions). Access to live video streams requires TLS-encrypted connections with mutual certificate authentication. The system must implement a privacy masking function that can blur designated zones — for example, windows of nearby buildings captured by external cameras — during recording and playback. Audit trails must log all access to video data, including timestamp, user identity, camera ID, accessed time range, and purpose of access, with logs retained for a minimum of 12 months.
| Parameter | Requirement | Typical Implementation |
|---|---|---|
| Compression standard | H.264 High Profile (mandatory), H.265 (recommended) | H.265 Main Profile for new builds |
| Recording mode | Continuous circular buffer + event-triggered retention | Continuous + pre/post event (30 s / 10 min) |
| Minimum retention | 72 hours continuous all cameras | 72 h to 720 h, depending on operator requirement |
| Storage redundancy | RAID 1 (mirrored) mandatory | RAID 1 with hot-swappable drives |
| Metadata | Timestamp, train ID, car, camera, event, speed | GPS-synchronized, stored in SQLite index |
| Encryption at rest | AES-256 mandatory | Hardware-accelerated AES-256-CBC |
| Power loss protection | Graceful shutdown, >= 30 s UPS | Supercapacitor-based UPS, auto file close |
Q1: Does IEC TS 62580-2 apply to all types of railway vehicles?
A: Yes, the standard applies to all rolling stock types including high-speed trains, mainline locomotives and passenger coaches, metro/subway vehicles, light rail/tram vehicles, and regional trains. However, the specific camera configuration and recording duration requirements may be adapted based on the vehicle type, operational environment, and local regulatory requirements. The standard provides a minimum baseline that operators can extend based on their specific security needs.
A: Yes, the standard applies to all rolling stock types including high-speed trains, mainline locomotives and passenger coaches, metro/subway vehicles, light rail/tram vehicles, and regional trains. However, the specific camera configuration and recording duration requirements may be adapted based on the vehicle type, operational environment, and local regulatory requirements. The standard provides a minimum baseline that operators can extend based on their specific security needs.
Q2: How is video data protected against tampering or unauthorized access?
A: The standard mandates AES-256 encryption for stored video data, role-based access control with graduated permissions, TLS-encrypted streaming with mutual certificate authentication, comprehensive audit logging of all access events, and cryptographic hash-based video integrity verification (SHA-256 hash chain) that can detect any modification of recorded video segments. For evidentiary purposes, a secure export function generates cryptographically signed video clips with a verifiable chain of custody.
A: The standard mandates AES-256 encryption for stored video data, role-based access control with graduated permissions, TLS-encrypted streaming with mutual certificate authentication, comprehensive audit logging of all access events, and cryptographic hash-based video integrity verification (SHA-256 hash chain) that can detect any modification of recorded video segments. For evidentiary purposes, a secure export function generates cryptographically signed video clips with a verifiable chain of custody.
Q3: What are the minimum video quality requirements for incident investigation?
A: For incident investigation, the standard requires that recorded video be of sufficient quality to enable person identification (facial features and clothing details) in the area of interest. This translates to a minimum resolution of 1080p at 25 fps for critical areas, with a minimum pixel density of 250 pixels per metre at the farthest point of the coverage zone. The standard recommends using the NIST PERF scale or equivalent for objective image quality assessment during system commissioning and periodic maintenance.
A: For incident investigation, the standard requires that recorded video be of sufficient quality to enable person identification (facial features and clothing details) in the area of interest. This translates to a minimum resolution of 1080p at 25 fps for critical areas, with a minimum pixel density of 250 pixels per metre at the farthest point of the coverage zone. The standard recommends using the NIST PERF scale or equivalent for objective image quality assessment during system commissioning and periodic maintenance.
Q4: Can railway CCTV systems use cloud storage instead of onboard recording?
A: While onboard recording is the primary storage architecture mandated by the standard, cloud or ground-based storage can be used as a supplementary tier. For real-time cloud streaming, the standard requires that the train-to-ground communication link provide a minimum sustained throughput appropriate to the number of cameras streamed simultaneously. Given the bandwidth limitations of current railway communication networks (typically 10-50 Mbps per train via 4G/5G), most operators use selective real-time streaming for alarm-triggered video while relying on onboard storage for bulk recording. The standard mandates that the onboard recorder always function as the primary recording source, with cloud storage serving as a backup and long-term archive tier.
A: While onboard recording is the primary storage architecture mandated by the standard, cloud or ground-based storage can be used as a supplementary tier. For real-time cloud streaming, the standard requires that the train-to-ground communication link provide a minimum sustained throughput appropriate to the number of cameras streamed simultaneously. Given the bandwidth limitations of current railway communication networks (typically 10-50 Mbps per train via 4G/5G), most operators use selective real-time streaming for alarm-triggered video while relying on onboard storage for bulk recording. The standard mandates that the onboard recorder always function as the primary recording source, with cloud storage serving as a backup and long-term archive tier.
