Stage 0: Mobilisation and calibration Before any seabed work starts, the survey spread has to be installed and proven. This stage is easy to rush, because it does not look like progress to the rest of the project. Typical tasks: Dimensional control of the vessel: measuring the offsets between the reference point, GNSS antennas, motion sensor, gyro, transducers and any crane or stinger reference points. Gyro and heading calibration alongside the quay, against known control. GNSS and acoustic positioning checks, including a USBL calibration if ROV or towed equipment will be tracked. Agreeing the survey procedures, reporting formats and the coordinate reference with the client before the first line is run. Everything later in the project rests on this stage. If an offset is wrong here, it will be wrong in every dataset until somebody notices. Stage 1: Route survey and pre-lay The route survey establishes the seabed conditions along the planned corridor: bathymetry, seabed features, obstructions and soil information from geophysical and geotechnical data. The pre-lay survey (or pre-installation survey) then confirms the corridor is still clear just before installation. Typical outputs: updated bathymetry, contact lists (debris, boulders, existing infrastructure), and confirmation that the route position list (RPL) is buildable. A good pre-lay report is short and decisive. The lay crew needs to know three things: is the corridor clear, where are the hazards, and has anything changed since the route survey. Long reports that hide the answer to those questions slow down the operation. Stage 2: Installation support During lay, survey is part of the control loop. Typical tasks include: Vessel and pipe/cable positioning relative to the RPL, shown to the lay crew in real time. Touchdown monitoring with ROV or an acoustic positioning system, so the engineers know where the product is actually landing, not where the model says. Tracking the cross-track deviation from the RPL and reporting excursions early. Calibrating the positioning and heading systems before the work starts. A calibration done after the problem is found is too late. Structures follow the same logic. When a jacket, subsea structure or spool is installed, survey provides the heading, position and level of the structure during set-down, and confirms it is within the installation tolerance before the crane releases. For spools and tie-ins, metrology measures the distance and angles between the two hubs so the spool can be fabricated to fit. A metrology error usually means a spool that does not fit, which is one of the most expensive surprises on a project. Stage 3: Post-lay and as-laid After installation, survey records what is actually on the seabed: Position of the pipeline or cable along its length, referenced to kilometre point (KP). Free spans, crossings, and proximity to other assets. Burial or cover, where required. The deliverables are usually alignment charts, event listings from the ROV video, and digital data in the format the client's integrity team uses. Eventing consistency matters here. If two shifts describe the same type of feature with different codes, the client's engineers will struggle to use the data years later. Stage 4: As-built The as-built dossier is the final record: final position, any rectification work, and the documents the client needs for integrity management over the asset's life. This is also where weak earlier records hurt. If the as-laid data is thin, the as-built is hard to defend. Remember who will use this dossier. It is not only the project team that is about to demobilise. It is the integrity engineer who opens it five or ten years later to plan an inspection, check a free span or design a new crossing. Clear coordinate information, consistent KP references and a short explanation of how each dataset was collected make that person's job possible. Missing metadata makes it guesswork. What links the stages One coordinate reference. Datum, projection and vertical reference must be fixed at the start and written down. Most expensive mistakes on construction projects are reference mismatches, not equipment failures. Traceability. Every number in the final report should trace back to raw data and a calibration record. Communication. Survey is a service to engineers, the lay crew and the client. A clean, timely, plain-language update is worth as much as a perfect dataset delivered late. Gaps I see often Calibration records stored on one laptop that leaves the vessel at demobilisation. KP references that change between revisions of the RPL without anyone telling the survey team. Daily reports that list what equipment was used but not what was found or decided. Last-minute requests for as-built data in a format nobody agreed at the start. Each of these is cheap to prevent early and expensive to fix at the end. Why this matters for your career If you can explain what each stage needs from the one before it, you are no longer "the person with the equipment". You are the person who prevents rework. That is the step from technician toward project and operations roles.