From the magazine – Accurate knowledge of a ship’s roll natural period is essential for safety, crew comfort, cargo security and avoiding extreme phenomena like parametric roll. MARIN has launched a free mobile app that helps with this.
This article originally appeared in SWZ|Maritime’s July-August 2026 issue. It was written by Rob Grin MSc, Senior Project Manager Seakeeping and Team Leader Special Vessels at MARIN.
Although several prediction methods are available (Grin, 2024), calculated periods can differ significantly from reality, with sometimes errors of several seconds. This can lead to incorrect safety assessments and operational decisions. This makes onboard measurement a good alternative. A dedicated roll period measurement app offers a simple, accessible, and reliable way to obtain accurate roll period estimates directly on board.
Measuring the roll period of a ship is not as straightforward as it may seem. A commonly used method is the “stopwatch” approach, in which a bridge officer typically times ten roll cycles and calculates the average period. While simple in principle, this method is prone to human error and requires notable roll motions. More importantly, the measured roll period is not always equal to the roll natural roll period of the ship.
In practice, discrepancies can be substantial. In the example in the figure below the text box below, the true natural period was 38.6 seconds, while repeated stopwatch measurements underestimated it by as much as 11.3 seconds (mean offset 3.8 seconds). Such deviations are unacceptable when the roll period is used for assessing risk on synchronous or parametric roll.
A more advanced technique involves spectral analysis of measured motions or accelerations. By applying a fast Fourier transform (FFT), the dominant period of the roll motion can be identified. Although more robust, accurate measurement depends on a sufficiently long measurement duration especially when roll motions are small. For instance, the FFT of the full time series in the example above has a deviation of only 0.1 seconds, whereas a fifteen-minute FFT (with root mean square (RMS) roll bigger than 0.5 degrees) can introduce offsets up to 1.2 seconds (mean offset 0.1 seconds).
Time series for a 15,000-TEU container ship
Model tests were conducted for the 15,000 TEU TopTier container vessel. This time series shows the vessel’s roll response in 3-metre stern-quartering seas, including one moderate parametric roll event after approximately two hours, reaching 10 degrees roll. The roll natural period in this loading condition was 38.6 seconds. Whenever the roll RMS (root mean square) exceeded 0.5 degrees, ten roll oscillations were counted, labelled I to XI. The largest deviation was observed in oscillation III, with a roll period of 27.3 seconds. The average offset was 3.8 seconds.
In addition, a direct fast Fourier transform (FFT) analysis was performed on eleven fifteen-minute segments, regardless of roll amplitude. For the first two fifteen-minute intervals, the roll period from the FFT were 18.3 seconds and 16.6 seconds, which is close to the wave encounter period, not the roll natural period. The remaining nine FFT results were within 1.2 seconds of the roll natural period, with a mean deviation of only 0.1 seconds.
The RollPeriod app concept
Until recently, dedicated equipment was required to measure motion or acceleration and perform FFT analysis. Today, every smartphone has these capabilities, which enabled us to develop an app that measures roll period. The idea is simple yet powerful: rather than relying on manual timing or specialised onboard systems, the app uses the phone itself as the measurement device.
Smartphones are equipped with accelerometers and gyroscopes capable of detecting orientation and motion with high precision. In the roll period app, these sensors are used to measure the effective gravity angle (EGA). This quantity combines the actual roll angle of the vessel with acceleration effects, similar to what a traditional inclinometer measures. By recording the longitudinal and transverse EGA at a sampling rate of 5 Hz (see the figure on the previous page), the pitch and roll periods can be determined through FFT analysis.
Also read: MARIN launches app for roll-period measurements
Measurement procedure
The RollPeriod app enables straightforward onboard use. A bridge officer only needs to place the phone horizontally, aligned parallel to the vessel’s centreline, and start the measurement. For reliable results, a measurement duration of fifteen minutes is recommended, but can be adjusted between two minutes and one hour. This fifteen-minute duration represents a balance between practicality and accuracy. Shorter measurements may already provide initial estimates, but longer records reduce uncertainty. It is recommended to repeat the measurement a few times at the start of the voyage.
When it consistently shows (nearly) the same roll period, the natural roll period is considered confirmed. Compared with manual timing with a stopwatch, the advantages are clear. The app eliminates human error and tedious waiting time. At the same time, it avoids the need for specialised instrumentation or post-processing software, making it accessible to a wide range of users.
Also read: How to act to prevent parametric rolling in following seas and avoid container loss
Future improvements
There are some challenges inherent to the measurement principle itself. The app determines the instantaneous roll period, which may slightly differ from the roll natural roll period depending on sea conditions and excitation. Detecting and correcting for this difference is an area of ongoing development. For instance, by better algorithms for distinguishing between encounter period and roll natural period.
An area for improvement is the import and export of the measured time series, calculated spectra and aggregated data. A real-time display of roll and pitch periods during measurements would also
help. It would give users immediate feedback while they wait for the full measurement.
Future versions could also incorporate the low fidelity roll risk estimator as developed within TopTier (available as download on TopTier | MARIN). A second feature could be the addition of an adaptive natural roll period prediction model that uses measurement data to refine its predictions. This would be beneficial for evaluating roll risk prior to departure.
Enhancing operational awareness and safety
The measurement of the roll natural period remains an important, but challenging task. Traditional measurement methods are either inaccurate or require dedicated equipment, while prediction models can deviate significantly from reality. The RollPeriod app represents a practical solution, combining the availability of smartphone technology with robust signal processing techniques.
By enabling quick and reliable measurements, the app has the potential to enhance operational awareness and safety at sea. Although further refinements may still be made, the app serves as a good example of turning theory into practice.
Also read: VIDEO: How to recognise and prevent excessive roll of container ships
REFERENCES
- R. Grin, “On empirical methods to predict the rolling period of ships”, 15th International Marine Design Conference, 2024
- TopTier | MARIN
The Rollperiod app is freely available from both the IOS App Store and Google Play Store.
Picture (top): More insight into a ship’s natural roll period can help prevent loss of cargo (photo by Flying Focus).
