Hello,
just want to ask about your strategies to monitor brightsign players.
I tested the prometheus-node-exporter, but I couldn't find a way to assure that the player shows content. A lot of metrics about CPU etc. but nothing else.
A simple ping is not helping, too to assure a playback is shown.
I was thinking about taking screenshots and have a script looking for pictures without content.
I want to monitor the players without using bsncloud.
What are your strategies?
Thanks,
lennard
Hello,
Prometheus node-exporter: Great for system metrics (CPU, memory), but it doesn’t report on what's being displayed.
Ping: Just confirms the device is online—not useful for confirming what's on screen.
Screenshot + image recognition: Solid approach, but needs some automation + logic to avoid false positives.
his is the most reliable, BSN.cloud-independent method.
How:
Use a device like a Raspberry Pi (or NUC) with an HDMI capture card connected to the BrightSign’s HDMI out (loopback).
Periodically grab a screenshot from the HDMI stream (e.g., using ffmpeg or v4l2).
Use a lightweight image recognition (or even color histogram comparison) to detect black/frozen/empty screens.
Optional: OCR with Tesseract to check for expected text overlays or timestamps in playback.
Benefits:
You see exactly what the display shows.
Can detect if the screen is black, frozen, or stuck.
Tools:
ffmpeg, v4l2-ctl, ImageMagick, Tesseract, OpenCV.
2. Custom Logging from BrightSign via Local Web Server / Syslog
BrightSign can be configured (via BrightAuthor or directly) to send logs or pings to a local server.
How:
Embed log messages or heartbeats inside the BrightAuthor project (e.g., every 60s).
Log could include:
“Now playing: Video A”
“Ad loop completed”
Timestamp, media name, duration, etc.
Push these logs to your local syslog server, or via HTTP to a small Flask/Node backend you control.
Benefits:
No need for HDMI capture.
Lower infrastructure cost.
Drawbacks:
EZPassNJ
Best Regards
Hi Lennard,
We use custom HTTP API checks to verify content playback (e.g., polling current_media). Alternatively, scrape BrightSign logs for errors or deploy a lightweight endpoint to confirm frame updates.
While BSN doesn't provide proof-of-play, it is useful and highly recommended.
Analyzing screenshots does sound like a solid approach. You could also check for the presence of a powered monitor on the HDMI output through EDID. Healthy screenshot alone doesn't mean that content is visible...
