4 reasons why autologging is not the solution to onboard data collection
Everybody agrees that in the future vessels will be equipped with sensors to automatically deliver all relevant data for vessel performance management. Looking at the technology status today, I still see a long way to go. Why?
1. Autologging does not deliver all data
Key information for performance management, such as the beginning / end of sea passage, a departure or the amount and type of cargo (e.g. number of reefer containers) are not available from autologging. Only the seafarer is able to tell the onboard data collection system that he left port or ended the river passage. Autologging data needs to be integrated with manual data, which requires an IT application. In contrast to the limited channels autologging systems provide, a lot of additional data is recorded manually in daily ship-to-shore reporting. This additional data can be easily used for further analysis, such as filters (i.e. all vessels in the same trade) but also in the sense of Big Data.
2. Autologging does not deliver significantly better data quality
As autologging requires a seamless integration from a physical sensor to an IT infrastructure and IT application, many things can go wrong. Sensors get stuck or break, and as nobody onboard looks at it), this can go unnoticed for some time. In addition, many shipboard sensors used are not meant for performance monitoring – they simply lack the precision for such a purpose. The IT interface gets interrupted or the application offers wrong or non-sufficient filters, averaging and plausibility checks. This is what a human trained brain does automatically, when looking e.g. at a power meter. In our projects, also when trialling autologging and manual reporting in parallel on the same vessels, we see the same quality levels of well reporting ship crews and good autologging systems. KPIs only differ by 1-2%.
3. Autologging does not deliver a new world of insights
Autologging data does have one significant advantage; it is more frequent than manual data. You could potentially get data every second. Today the rate it is sent onshore and processed is typically 15 min data. There are performance-relevant KPIs that you can only see with high frequency data, such as rudder movements or engine overload. But there are not many extra indicators and insights gained through high frequency data, which means we are not opening up a new world of insights through autologging.
4. Autologging needs considerable project effort to make (and keep) it work
On top of the price, the effort to get everything in place, set up, installed and running is significant. Especially when completed for a vessel in operation (i.e. not a newbuilding). A lot of preparation time to match the vessels hardware with the autologging system is needed. A service technician needs to sail with the vessel for a few days and the system and sensors need to be kept alive, so repair and maintenance effort is needed. That is why you hardly ever see large fleet roll outs of 50+ vessels with autologging systems. Very often it is 5 vessels here, 10 vessels there.