How the GMSL Health Monitoring and Error Recovery feature in e-con Systems’ STURDeCAM Series Enhances Autonomous Mobility

GMSL technology is a high-speed and power-efficient video link commonly used in advanced driver assistance systems (ADAS), smart traffic, and robotic applications. It operates by employing serializers to convert data into a serial stream on the transmitter side and deserializers to convert it back into a parallel word for processing on the receiver side.

However, the increasing need for continuous and secure data transmission across the GMSL link in diverse embedded vision applications has become paramount. Addressing this demand, e-con Systems offers the STURDeCAM series, which consists of advanced cameras equipped with a robust GMSL health monitoring and error recovery feature — built on our newly granted patent, “System and Method for Recovering Camera Streaming Failures”. These ensure the safety and reliability of data transmission over GMSL, preventing system failures and enabling timely diagnostics for enhanced security in applications. This GMSL health monitoring and error recovery feature is currently supported on STURDeCAM cameras with on-board ISP, where the camera driver has direct access to the ISP for diagnostics and recovery.

Let’s understand the significance of the GMSL Health Monitoring and Error Recovery feature in e-con Systems’ STURDeCAM series and see how this enhances the resilience of data streaming.

Role of GMSL Health Monitoring in Ensuring Data Transmission

The STURDeCAM series by e-con Systems comprises several GMSL2 cameras with on-board ISP that implement robust Health Monitoring and Error Recovery capabilities that diligently oversee the camera streaming state. GMSL health monitoring plays a key role in ensuring robust camera streaming by comprehensively monitoring the sensor, Microcontroller Unit (MCU), and GMSL link.

This continuous surveillance minimizes data transmission errors, guaranteeing reliable control signal reception and accurate data processing. Automated diagnostics can conduct real-time evaluations of data integrity, signal strength, and link stability, creating a robust fail-safe mechanism to safeguard the streaming process.

How does the GMSL Health Monitoring and Error Recovery Feature in the STURDeCAM series work?

In instances where errors or failures arise during camera streaming, users often turn to manual interventions to address these issues. The following steps are commonly performed:

1. Restart the application
2. Stop the application – reload the driver – start the application
3. Reboot the system – load the driver – start the application

These manual interventions are a traditional approach to troubleshoot and resume the camera streaming functionality. However, the GMSL Health Monitoring and Error Recovery feature in the STURDeCAM cameras eliminates the need for manual interventions, ensuring a more efficient and seamless camera streaming experience. Now, let’s delve into the operational details of this innovative feature and understand how it enhances the reliability of the camera.

First, check out how this feature works and how it eliminates the above-mentioned challenges.

Figure 1: GMSL Health Monitoring and Error Recovery Working Process

Here’s an overview of the process:

Sensor Driver Vigilance

The camera streaming state is continuously monitored by the sensor driver to ensure seamless data flow.

Event Generation and Notification

Any interruption in camera streaming triggers an event, promptly notifying the camera driver of a potential issue.

Error Identification and Recovery Process

The camera driver then initiates a thorough investigation, checking the deserializer, serializer, microcontroller, sensor, ISP status, and more to identify the root cause of the interruption.

The steps involved in the recovery process encompass a thorough examination of each component:

1. Finding the Erroneous Camera
2. Serializer and Deserializer Error Handling
3. MCU Error Handling
4. ISP Error Handling

Finding the Erroneous Camera: The recovery process initiates upon the detection of uncorrectable errors in a specific camera. This prompts the activation of a monitoring thread responsible for identifying the problematic camera. Leveraging the exported device name, this active thread precisely identifies the source of the error.

Serializer and Deserializer Error Handling: The process proceeds with a detailed examination of the serializer and deserializer crucial for data transmission. In general, the image sensor’s output is received at a serializer that converts it into serial data, transmitted via coaxial cable to the deserializer and deserialized for platform processing.

Upon identifying the error camera, the sensor driver verifies the link lock status of the serializer and deserializer. Additionally, a detailed PCLK (Pixel Clock) detection process is executed specifically for serializer validation and transmission of valid data confirmation.

MCU Error Handling: The system verifies the microcontroller unit’s (MCU) availability and functionality by retrieving its firmware version. Upon successful verification, the system initiates ISP reinitialization and stream configuration.

ISP Error Handling: The process proceeds with the validation of ISP functionality through chip ID reading, streaming mode check, and frame count verification. This confirms ISP communication, verifies the streaming state, and ensures frame output integrity for precise and accurate data transmission.

Logging, Recovery, and User Control: The recovery process logs the issue’s root cause to the DMESG, enabling post-analysis and system enhancement. Users can retrieve device states through the SYSFS entry at any time.

Moreover, the user has the flexibility to enable/disable Error Recovery through the SYSFS entry, with the feature being enabled by default. The SYSFS entry also allows users to get real-time status updates regarding the health monitoring of a specific camera.

Ensuring Reliability with GMSL Health Monitoring and Recovery across Autonomous Mobility Platforms

Autonomous mobility, spanning ADAS, Delivery Robots, Autonomous Mobile Robots (AMR), and Autonomous Agriculture vehicles, relies heavily on real-time data for seamless operations. The STURDeCAM camera series, equipped with a robust GMSL-based health monitoring and recovery system, ensures reliability across these applications:

Real-Time Data Integrity for ADAS

GMSL link monitoring ensures a continuous flow of accurate and secure data, crucial for real-time decision-making in ADAS functionalities such as lane departure warning, adaptive cruise control, collision avoidance, and vehicle safety enhancement.

Precision in Delivery Robot Navigation

Continuous and accurate transmission of visual data aids delivery robots in navigating various terrains and environments with precision, ensuring accurate route planning and obstacle avoidance.

Support across Autonomous Systems

The GMSL health monitoring and recovery features support resilient data streaming for Autonomous Mobile Robots (AMRs) in logistics and warehouse management. It augments visual data accuracy in autonomous vehicles used in precision farming and facilitates parking lot management.

The integration of GMSL health monitoring and error recovery features in e-con Systems’ STURDeCAM series sets a new benchmark for reliability in autonomous mobility. By ensuring uninterrupted data transmission, dependable control signal reception, and robust fail-safe mechanisms through automated diagnostics, this system paves the way for safer, more reliable autonomous mobility solutions.

Complementing this groundbreaking feature, e-con Systems’ cameras also incorporate innovative technologies such as sub-pixel technology, Low-light 120dB HDR without Motion artifacts, LED Flicker Mitigation (LFM), and an IP69K-rated design. These additional cutting-edge functionalities further elevate its performance, ensuring optimal functionality even in challenging environmental conditions.

Explore our STURDeCAM series

Visit our Camera Selector page for a comprehensive view of our products.

If you are looking for more information on our STURDeCAM series or need any help integrating any other camera solutions, please write to us at camerasolutions@e-consystems.com.