Give each occurrence its own time, full message and step name. Keep the tray identity with it. Pick the one with the clearest matching trace, rather than combining the best-looking pieces from both. That would manufacture a third incident nobody observed.
Include what happened just before that pause. Did the operator move the tray, change the job, or simply watch the next step wait? Sequence matters more than another minute of a stationary arm.
Has the seller received that specific set, Alex? I'd ask them to answer against its incident time. Otherwise you may get a diagnosis based on the first clip while you're talking about the second.
Get maintenance to supply that from the matching capture. Also describe the tray's physical location in the photograph. A signal reading and a tray that looks seated from one angle still need interpreting together.
Keep the first unexplained occurrence open too. This finding may explain the second one, but you said the first lacks the message. I'd rather leave that limit clear than claim the whole intermittent problem is solved.