Polling and/or Tracing

In the realm of data acquisition, two primary methodologies reign supreme: polling and tracing. These approaches serve as the backbone for capturing crucial data from sensors or measurement devices, each offering distinct advantages and tailored use cases. The decision between polling and tracing hinges upon the unique demands of the application at hand. Leveraging ATI’s VISION software in conjunction with high-speed ECU serial interfaces, significant strides can be made in obtaining large data item counts efficiently and effectively.

Polling involves actively requesting data from a sensor or measurement device at predetermined intervals. The data acquisition system sends requests to the sensors, asking for the current readings. The system then waits for a response before making the next request. Some advantages are - control over data acquisition frequency: you can precisely control when data is collected; lower data volume: data is only collected when needed, which can be beneficial for applications with limited storage or bandwidth and real-time responsiveness. There is immediate access to the latest data, ensuring timely adjustments.

Tracing, also known as streaming or continuous acquisition, takes a different approach. Instead of actively seeking out data, tracing involves continuously recording information from sensors, creating an unbroken flow of measurements without activity requesting it at every instance. As such, tracing is well-suited for applications where continuous, real-time data is essential. It enables a higher data throughput which is beneficial for applications requiring a high sample rate. Finally, it requires a simplified control logic because it eliminates the need for frequent polling requests, streamlining operations.

Choosing between polling and tracing can depend on several factors. Firstly, specific needs of your application. If you need real-time responsiveness and control over data acquisition frequency, polling might be more suitable. If continuous monitoring is crucial, and you require a high sample rate, tracing may be the better choice.

Secondly, available resources, such as storage capacity and bandwidth. Polling can be more efficient in terms of data volume, making it suitable for applications with limited resources.

Lastly, latency considerations. If low latency is crucial, polling can provide immediate access to the latest data. Tracing might introduce some latency, depending on the buffering and transmission mechanisms.

In practice, the choice between polling and tracing depends on the specific requirements of the application. Let’s breakdown some potential real-world user-cases in detail:

Polling:

a.  Scenario: Heating and cooling - temperature sensors used for battery systems to monitor temperatures at various points.

b.  Reason for Polling: Polling can be effective when users require periodic updates on temperature changes. It allows users to control the frequency of temperature data collection, conserving resources and ensuring real-time responsiveness when adjustments are needed.

a.  Scenario: In battery-operated devices, monitoring the battery voltage is critical to prevent unexpected system shutdowns.

b.  Reason for Polling: Polling allows the system to periodically check the battery voltage without continuously draining the battery. This helps manage power consumption effectively.

Tracing:

a.  Scenario: Analyzing high-frequency signals, such as those in signal processing applications.

b.  Reason for Tracing: Tracing is suitable when users require a continuous stream of data to capture rapid changes in signals. It ensures that users don’t miss critical data points and allows for high sample rates.

a.  Scenario: In automotive testing, monitoring various parameters (speed, acceleration, etc.) during a test drive.

b.  Reasons for Tracing: Tracing is beneficial when a constant flow of data is required to analyze the vehicle’s performance throughout the entire testing process. It provides a comprehensive data set for in-depth analysis.