Google Translate says: - they don't actually know what triggered the switch into manual mode "RWS - Programs, Projects and Maintenance (hereinafter: PPO) has

Google Translate says: - they don't actually know what triggered the switch into manual mode "RWS - Programs, Projects and Maintenance (hereinafter: PPO) has conducted research done, and had done, to the technical cause for the switch to manual mode. PPO has not yet been able to determine the cause. The reason for this is that the logging circuits in the system approximately 1 x per second registers. Switchings can take place in approximately 200 ms according to PPO. This makes it impossible to determine which circuits led to the failure." - their SCADA in general seems poorly maintained: "According to the maintenance contractor, there are approximately 60 faults active in SCADA in normal operation" "In the past, a software issue was found in the vibration measurement system of the pumping station pumps. This caused a technical safety function to be disabled." "During work on the installation of pump 7 in the pumping station, various software errors were found by the maintenance contractor." - and operation degenerated into treating it as arcane, poorly understood black box: "In the event of failure of regular operation, the sluice and pumping station complex can be operated from Schellingwoude using the emergency panel. It is important that this system is used in an emergency. During the incident, this was not used, because operators did not trust the functioning of the panel and were not well-acquainted with the operation of the panel." - and of course the UX (or Operator Experience? OX?) is poor too: "In SCADA, the sluice gates are light grey instead of dark grey when the sluice complex is switched to manual operation. However, the light grey sluice gates also occur in various phases of regular operations." Obvious caveats regarding correctness of Google Translate apply; I do not speak Dutch. Most of this is from page 14 (or page 11 of the "INTERGO" sub-embedded report). And I'm focusing on the negative here — Amsterdam didn't flood, so there was (even if just barely) enough "buffer" in their processes to prevent the worst.