When the 400-metre-long container giant “Ever Given” obstructed the Suez Canal in March 2021, it froze 12% of world trade for nearly a week. The episode exposed how little room for error exists in the confined waterways that carry the world’s largest ships, and how difficult it is for a human crew to react quickly enough when tonnes of steel start drifting. As digitalisation and smart shipping forge new pathways to enhance maritime transportation, one question generates utmost curiosity: Can autonomous shipping prevent another Ever Given incident? Our research [1] suggests that equipping ships with a digital co-pilot may be an effective solution for making navigation safer and more efficient.
The grounded container ship ‘Ever Given’ in the Suez Canal (2021) (Picture credits: CNN)
Inside the autonomous wheelhouse
In our recent collaboration within the AUTOBarge project [1], we designed a nonlinear model-predictive controller (NMPC) tailor-made for narrow canals and rivers. We employed a prediction model that explicitly includes shallow-water effect, asymmetric current profiles and most importantly, the bank-effect forces suspected in the Suez Canal grounding. By forecasting how these forces will grow, the NMPC begins gently applying a counter-rudder action, keeping the hull centred with the least effort. This is essential when a single degree of yaw can translate into meters of lateral drift in a tight fairway.
In simulated trials, the controller drove a 100-m pusher-barge convoy through three classic choke-points: a straight 100 m-wide canal, a T-junction and a winding river bend. Whereas the conventional PID controller lets the convoy wander significantly off course, the predictive controller cuts the maximum cross-track error to well under strict tolerances across all scenarios. At a tricky T-junction, the NMPC steered through the 90° turn with less than 5 m error, while the PID solution strayed over 20 m before recovering. Those numbers matter: the Ever Given’s first contact with the eastern bank came after a drift of only a few meters.
Accidents do not necessarily need challenging environmental conditions to occur, as evident by the recent grounding incident in Trondheim, where a container ship crashed into the front garden of a resident. (Picture credits: Richard Sagen, Adressa.no)
Efficiency as a bonus, not an afterthought
Because the NMPC chooses smoother rudder angles, it spends far less hydraulic energy steering than a reactive PID controller. In straight-canal tests, the predictive approach cuts rudder effort by an order of magnitude. On winding river-bend scenarios, it shaved almost 200s off a 20-minute run, an 11 % time saving that translates directly into lower fuel consumption at constant engine RPM. We also propose key performance indices for safety, efficiency and robustness, so that skippers and regulators can quantify the gains.
Why this could have changed the Suez story
AIS video feed of the ‘Ever Given’ running aground in the Suez Canal (Source: Maritime Casualty Specialists, Italy).
Bank-effect forces, the very phenomenon that slewed the Ever Given off course, are embedded in the NMPC’s hydrodynamic model. The controller “sees” the sideways suction well before it becomes visible on the bridge instruments and begins counter-steering gently instead of waiting for an alarm. Simulations show that when currents reverse, analogous to the cross-winds blowing in the canal on 23 March 2021, the predictive controller still keeps the vessel within the canal’s half-width margin. In short, autonomy can recover the seconds that human reaction loses.
Why the next Ever Given-size casualty is avoidable
Autonomous and smart shipping does not have to mean crewless. It means providing captains with a “digital” co-pilot that can detect potential risks before they escalate. The evidence from the performed studies is clear: ships that can anticipate rather than react prevent the metre-scale drifts that lead to groundings like the Suez blockage. Retrofitting a predictive “guardian” to existing helms could be the fastest way to ensure the next megaship entering the Suez, or any other narrow sea-lane or canal, never repeats the Ever Given’s wrong-way turn.
An article by Abhishek Dhyani
References
[1] Zhang, C., Dhyani, A., Ringsberg, J. W., Thies, F., Negenborn, R. R., & Reppa, V. (2025). Nonlinear model predictive control for path following of autonomous inland vessels in confined waterways. Ocean Engineering, 334, 121592. https://doi.org/10.1016/j.oceaneng.2025.121592