← Back to Archive

Nothing underwater travels in a straight line

Currents, trim and the constant correction a vehicle needs.

Nothing underwater travels in a straight line unaided because currents, however gentle, are constantly pushing a vehicle sideways off whatever heading it was given, and a vehicle's own trim, its tendency to drift or tilt if even slightly unbalanced, adds a second constant force nudging it further off course, so holding a straight line underwater is never a passive default the way coasting down a quiet, well-marked road can feel, it is an active correction repeated continuously for as long as the vehicle is moving.

The short version

A current is simply the surrounding water itself moving, and a vehicle sitting inside a moving current is being carried along with it in exactly the same way a leaf dropped onto a stream drifts downstream regardless of anything the leaf itself is doing, except that the vehicle also has its own thrust pushing it somewhere else at the same time, and the two effects add together rather than one simply overriding the other. Trim adds a second, quieter push of its own, since even a hull built and loaded with real care rarely balances its weight and buoyancy perfectly, and the smallest residual imbalance produces a slow, steady tendency to roll, pitch, or yaw away from a straight line even in still water with no current at all. Neither effect announces itself the way a pothole or a crosswind does on a road, since there are no painted lines, no fixed roadside markers, and often no visible reference of any kind underwater, which means a vehicle can drift steadily off course for a long time before anything on board even registers that it has happened. A car left entirely alone on a flat, straight, windless road will actually hold something close to a straight line for a surprisingly long stretch, since gravity and tyre friction are quietly doing useful work the whole time without being asked, and it is exactly that quiet, free assistance which has no underwater equivalent at all.

The river-crossing comparison

Paddling a canoe straight across a flowing river by simply pointing the bow at a landmark on the far bank and paddling directly toward it reliably lands well downstream of that landmark, since the current has been quietly carrying the canoe sideways the entire time it was moving forward, and only actively angling the bow upstream, adjusting continuously as the crossing proceeds, keeps the canoe tracking the straight line the paddler actually intended. An underwater vehicle crossing any current is solving exactly the same problem, needing to point somewhat into the current rather than directly at its destination, and needing that correction recalculated continuously rather than set once at the start, since the current itself can shift in strength or direction over the course of a single crossing in a way a fixed upstream angle chosen at the outset would not account for. A paddler who does correct constantly, reading the far bank's slow sideways creep and adjusting stroke by stroke, arrives close to the intended landing point without ever needing to know the current's exact speed, which is the same trick an underwater vehicle's control system is really performing, correcting against the drift it can actually observe rather than trying to predict a current it can only ever measure indirectly.

The one number worth remembering

A heading error of only a few degrees, held uncorrected over a run of several hundred metres, can leave a vehicle tens of metres off its intended track by the time it arrives, an error easy to underestimate because a few degrees looks trivial on a compass but compounds steadily with every metre travelled rather than staying fixed. That is precisely why underwater navigation treats a small, early heading error as a problem worth catching immediately rather than one worth ignoring until it becomes visible, since by the time an uncorrected drift is large enough to notice without instruments, the distance already lost to it is usually far larger than the small correction that would have prevented it from the very start. A current pushing steadily from one side has exactly the same compounding character as the heading error itself, since even a gentle sideways push, sustained for the whole length of a journey rather than corrected along the way, moves a vehicle a genuinely large distance off its intended line by the time the journey is over.

What follows from this

Because currents and trim are never truly absent, underwater vehicles are built around continuous correction rather than a heading set once and left alone, using onboard sensors to measure the vehicle's actual heading, depth, and attitude many times a second and feeding the difference between that measurement and the intended course into a control theory loop that nudges the thrusters accordingly. That loop runs constantly for the entire length of a journey rather than only when something visibly goes wrong, since the whole point of the correction is to prevent an error from ever becoming visible in the first place, catching a drift of a fraction of a degree long before it has any chance to compound into the kind of gap the previous section describes. The thrusters doing that correcting are the same ones described earlier in this set, placed at the hull's extremities for exactly the turning leverage this constant, low-level nudging depends on, so a vehicle built with thrusters clustered too close to its own centre would be paying a real, ongoing cost in how tightly it can actually hold a line, quite separately from how well it can turn on command.

What this does not explain

None of this explains how a vehicle actually discovers that it has drifted in the first place, since correcting a known error and detecting an unknown one are genuinely separate problems, and underwater there is no equivalent of glancing at painted lane markings or checking a satellite signal to confirm a position directly. That detection problem leans on the one sense that reliably reaches any real distance underwater at all, which the final article in this set turns to directly.

More on Propulsion in water