View the problem of reaching the rocket as similar to locating the head of a CNC or 3D printer. You need XY positioning. Put the gripper on a bridge that spans the ship. The gripper moves along that bridge for movement in one direction. The bridge itself moves along tracks in the sides of the ship, providing the other dimension of movement. The mechanism can be doubled up to grab from both sides simultaneously, possibly eliminating the need for an active gripper.
Another way is an arm. An ordinary heavy-equipment arm, as used for digging up roads, could reach out and pin down a leg of the rocket. It need not be automatic; a human could remotely operate it. The vertical dimension isn't really needed, so sliding the arm is also an option.
Another way is to flood the deck with adhesive, possibly something like molten solder or polyurethane foam.
Another way is to pre-place grippers all over the deck. Make a deck of grippers. They just grip; they do not need to move around on the deck. It can be mechanical, electromagnetic, or vacuum.
None of which are trivial in terms of expense or engineering.
View the problem of reaching the rocket as similar to locating the head of a CNC or 3D printer. You need XY positioning.
To reach all the legs, you'd need two. As these are the dimension of the entire landing pad, they'd cost more. They're essentially gantries. Also, they'd still need to be engineered to withstand and remain operational in heavy seas. Would they be running on hydraulics? What of the effects of the rocket blast. What happens if there is a hydraulic oil leak?
Another way is an arm. An ordinary heavy-equipment arm, as used for digging up roads, could reach out and pin down a leg of the rocket.
Those would certainly be hydraulic, so the same objections above apply. This is a good use of existing commodity equipment -- except that it would have to be hardened against the sea and the rocket blast as well, so it would still be an expensive specialty item.
It need not be automatic; a human could remotely operate it.
Keeping telemetry going while the rocket is landing is still something they haven't solved.
Another way is to flood the deck with adhesive, possibly something like molten solder or polyurethane foam.
It's thinking outside the box, but this isn't going to work 100% in heavy seas. I suspect that any adhesive that would work in heavy seas in the necessary time frames would be expensive to remove afterwards.
Another way is to pre-place grippers all over the deck. Make a deck of grippers.
How about make the deck out of slotted rails? Make gripper robots which are designed to move on these rails or between rails? Also the slotted rail system can be made into a 2D grid.
The rocket blast isn't anything special for heavy equipment that is some distance away. Note the containers sitting on the ship. They are fine.
Grabbing can happen a minute after the engines stop, so telemetry isn't a problem.