Technical Forms Surveillance - Applied AI Foundation

A vessel files dozens of technical forms a month — ME performance, AE performance, deflection, bearing clearance, scrape-down, scavenge inspection, lube-oil consumption, lube-oil shore analysis, lube-oil condition, paint inventory, chemical inventory, walkie-talkie inventory. Two questions matter:

Are they being submitted on time? (compliance)

What do the numbers in them actually tell us? (analysis)

Most forms-management systems answer only the first. The pipeline below answers both, on every monthly submission, automatically.

Where the data comes from

Source	What it provides
Vessel ERP (forms module)	Form register, mandatory list, submitted dates, due dates
Vessel-filed forms	Structured form data — engine performance, lube-oil consumption, scavenge, paint & chemical inventory, walkie-talkie inventory, etc.
DOC compliance master	Document-of-Compliance form requirements per vessel type and flag
Email correspondence	Form-chase threads from shore to vessel

Forms are filed by the vessel’s officers via the ERP’s forms module. The pipeline reads each filed form’s structured payload (it does not re-parse the PDF) and runs the analytical layer per form. Submission status comes from the same ERP register.

Compliance — the submission view

For every required form, the pipeline tracks the last submitted date, the next-due date, and the lateness:

\Delta_\text{days} = D_\text{next\_due} - D_\text{today}

Submission rate per vessel per period:

\text{Submission rate \%} = \frac{N_\text{submitted on time}}{N_\text{mandatory}} \times 100

Three submission views:

Previous month
Previous 12 months
DOC-wise

The most recent closed month — what got submitted on time, what was late, what’s missing entirely. The view a Technical Superintendent reads first.

The lateness distribution is more useful than the binary on-time count:

Window	Bucket
0–7 days late	Acceptable slip
8–30 days late	Concerning
31+ days late	Compliance gap — escalate
Not submitted	Black box — every analysis below depends on this form

Analysis — what each form actually tells us

The compliance view is necessary; the analytical view is where the value is. Every monthly form contains numbers that reveal something about the vessel. The forms-processing templates run an analysis on each form as soon as it’s filed, and route findings into the relevant pipeline.

ME performance form

Running hours, load, SFOC, Pmax, Pcomp, exhaust pressure, exhaust temperature, feed rate. Routes findings to the ME performance pipeline for cylinder uniformity and SFOC variance analysis.

ME bearing clearance

Main bearings, bottom-end bearings, crosshead bearings, thrust-pad clearance — measured against the maker’s wear-down curve:

\text{Wear \%} = \frac{C_\text{measured} - C_\text{new}}{C_\text{condemn} - C_\text{new}} \times 100

Wear %	Status
$\leq 50\%$	OK
$50\%$ to $80\%$	Monitor
$80\%$ to $100\%$	Plan replacement
$> 100\%$	Condemn — must be replaced

ME scrape-down

Iron content from cylinder oil scrape-down samples, residual BN variation, stuffing-box leakage status. Iron content is a leading indicator of liner / ring-pack wear:

Iron content	Interpretation
$< 100$ ppm	Normal break-in
$100$ to $250$ ppm	Active wear, monitor
$250$ to $500$ ppm	Accelerated wear, investigate
$> 500$ ppm	Severe — engine inspection required

Stuffing-box leakage logged here cross-references the lube-oil consumption variance form.

ME scavenge inspection

Feed rate, lubrication status, deposit formation, coating thickness, axial clearance per unit. The scavenge inspection is one of the few times the engineers actually see inside the engine, so the report quality matters.

AE performance form

Same shape as ME but per AE. Routes to AE performance pipeline.

Lube oil condition

TBN, water content, viscosity, renewal status for ME / stern tube / AEs. Routes to lube oil pipeline.

Deflection forms (ME and AE)

Deflection readings per crankshaft web — indicator of bearing wear and crankshaft alignment. Out-of-tolerance deflection is a major-overhaul flag.

ICCP performance

Aft, midship, forward current outputs and reference voltages from the impressed-current cathodic protection system. An ICCP that’s drawing too much current is a paint condition issue; one drawing too little is an electrical issue.

Inventory forms

Paint inventory — ROB vs reorder threshold, drydock prep adequacy
Chemical inventory — water-treatment chemicals, cleaning chemicals; consumption vs spec
Walkie-talkie inventory — count, condition, charging — a small form but PSC inspectors check it
Lube oil monthly — ROB, consumption, FOC, renewal status (cross-reference with lube-oil pipeline)

Motor condition analysis

Megger readings and current draw on critical motors — bilge pump, fire pump, GS pump, AE starter motors. Falling insulation resistance is a leading indicator of motor failure.

Month-end consolidated report

The roll-up: every monthly form analyzed, with the worst findings surfaced first. The view a TSI reads on the first of every month.

Implementation reference

The submission status logic is structurally identical across the three submission views — only the time window differs. Condensed:

def compute_submission_status(forms, today):
    rows = []
    for form in forms:
        last_submitted = parse_date(form.get("submittedDate"))
        next_due       = parse_date(form.get("nextDueDate"))

        if last_submitted is None:
            status = {"text": "Never submitted", "color": "red"}
        elif next_due is None:
            status = {"text": "In Order", "color": "green"}
        else:
            days_diff = (next_due - today).days
            if days_diff < 0:
                if abs(days_diff) <= 7:
                    status = {"text": f"Late {abs(days_diff)} days", "color": "orange"}
                elif abs(days_diff) <= 30:
                    status = {"text": f"Late {abs(days_diff)} days", "color": "red"}
                else:
                    status = {"text": "Compliance gap", "color": "red"}
            elif days_diff <= 14:
                status = {"text": f"Due in {days_diff} days", "color": "orange"}
            else:
                status = {"text": "In Order", "color": "green"}

        rows.append({
            "form":           form["name"],
            "last_submitted": last_submitted,
            "next_due":       next_due,
            "status":         status,
        })
    return rows

The analytical templates all follow the same shape: read the form’s structured data, extract the parameters, classify each parameter against a maker’s spec or fleet-norm threshold, return both the per-parameter verdict and the aggregate form verdict.

Repeat-delay detection

A form that’s late once is a slip; a form that’s been late three months in a row is a process problem. The pipeline groups late submissions by form and surfaces forms that have failed three or more consecutive periods:

RECURRING LATE SUBMISSIONS

ME deflection           →  Late 4 of last 6 months
Lube oil shore analysis →  Late 5 of last 6 months
Paint inventory         →  Late 3 of last 6 months

Recurring late submissions usually trace back to one of:

Owner of the form rotated off vessel and handover incomplete
Form has been moved into the master compliance system but the engineers don’t know
Form depends on a sample / measurement that itself is overdue
Form has been deprecated but not removed from the mandatory list

The first three are recoverable with one conversation; the fourth is a maintenance gap in the master compliance system.

Worked example

MV POSUN, end-of-April submission review:

Form	Last submitted	Status	Notes
ME performance	2026-04-05	In Order	—
AE performance	2026-04-08	In Order	—
AE LO consumption	2026-03-12	Late 18 days	Recurring — late 3 of 6
LO shore analysis	2026-02-25	Compliance gap	35 days late
ME bearing clearance	2026-04-15	In Order	Crosshead 4 wear at 78% — monitor
ME scrape-down	2026-04-15	In Order	Iron 165 ppm — normal
Paint inventory	2026-03-30	Late 8 days	Recurring — late 3 of 6
Walkie-talkie inventory	2026-04-12	In Order	All 8 units OK

Submission rate: 84% — below the 95% threshold, flagged. Recurring lates: AE LO consumption, LO shore analysis, paint inventory — all three are likely linked to the same officer change. Analytical findings routed to other pipelines:

ME crosshead 4 bearing at 78% wear → routed to ME performance and PMS for replacement planning.
Paint inventory recurring lateness → flagged to chief officer for handover audit.
LO shore analysis 35 days late → blind spot on lube oil for the past month, escalated.

What the senior review contains

Submission rate — overall %, lateness distribution, gap count.
Recurring late forms — the pattern view, with likely cause.
Compliance gaps — forms more than 30 days late or never submitted.
Analytical findings by form — every analyzed form with its verdict, routed to the appropriate domain.
Cross-domain notes — where a form’s content matters for another pipeline (e.g. wear-down readings for PMS planning).
Recommendations — split between submission-process fixes and analytical action items.
Escalation decision — to whom, and why.

Escalation triggers

Trigger	Severity
Submission rate below 90%	HIGH
Form 31+ days late	CRITICAL
Repeat delay across 3+ consecutive periods	HIGH
Safety-critical form (lifesaving, firefighting) overdue	CRITICAL
Form analytical finding outside spec (e.g. bearing condemn)	per finding
Compliance gap on lube-oil consumption or shore-analysis forms	HIGH

Why both views matter

A vessel can have 100% submission rate and still have a wear-down problem buried in the deflection report nobody read. A vessel can have low submission rate and be operating fine — until the audit lands. The pipeline reports both because they’re independent failure modes; you can fail one and pass the other, and the actions are different.

The forms compliance number is what auditors care about. The analytical findings are what keeps the engine running. They are not the same problem and shouldn’t be conflated.

References

Forms-submission templates

Submission status views — previous month, previous 12 months, and DOC-wise rollup.

Templates: forms-processing

15 templates running per-form analysis on ME / AE performance, bearing clearance, scrape-down, scavenge inspection, deflection, lube oil, ICCP, motor condition, paint & chemical inventory, walkie-talkie inventory, and the month-end consolidated report.

Related: ME / AE performance

ME and AE performance forms feed directly into the engine pipelines.

Related: Lube oil

Lube-oil consumption and shore-analysis forms are the source data for the lube oil pipeline.

Documentation Index

​Where the data comes from

​Compliance — the submission view

​Analysis — what each form actually tells us

​ME performance form

​ME bearing clearance

​ME scrape-down

​ME scavenge inspection

​AE performance form

​Lube oil condition

​Deflection forms (ME and AE)

​ICCP performance

​Inventory forms

​Motor condition analysis

​Month-end consolidated report

​Implementation reference

​Repeat-delay detection

​Worked example

​What the senior review contains

​Escalation triggers

​Why both views matter

​References

Forms-submission templates

Templates: forms-processing

Related: ME / AE performance

Related: Lube oil

Where the data comes from

Compliance — the submission view

Analysis — what each form actually tells us

ME performance form

ME bearing clearance

ME scrape-down

ME scavenge inspection

AE performance form

Lube oil condition

Deflection forms (ME and AE)

ICCP performance

Inventory forms

Motor condition analysis

Month-end consolidated report

Implementation reference

Repeat-delay detection

Worked example

What the senior review contains

Escalation triggers

Why both views matter

References