Engineering Cost Estimation¶
This page records the reasoning behind the man-hour factors entered in hote's Engineering Cost Estimator (Settings → Engineering Cost Estimator → Manage Factors), so the numbers can be reviewed and updated by anyone on the team, not just whoever typed them in.
How the estimator applies these factors¶
Each factor row defines hours across seven phases (Engineering, PM,
Documentation, FAT, Assembly, Training, Test & Commissioning) for one
basis unit (e.g. 1 x ea). When building an actual estimate for a
proposal, you pick the relevant factor row(s) and enter a quantity —
that quantity multiplies every phase-hour on the row (an NSR checkbox
applies an additional 1.5× for non-standard requirements).
Each row also has a one-time overhead (hours and/or a flat cost) that is added once per row selected, regardless of quantity — this is what lets a single row combine a per-unit rate with a fixed setup cost without either one distorting the other as quantity changes.
The basis unit field is editable — pick from the established shorthand
list (ea, set, lot, hr, md, etc.) when one genuinely fits, or
type a more descriptive unit like camera or speaker when that makes
the row clearer at a glance. Clarity wins over forcing everything into
the generic list.
Global settings (man-hour rate, working hours/day) are set once in Manage Factors and apply to every estimate.
CCTV Coverage Study (JVSG)¶
A CCTV coverage study is the JVSG modelling exercise used to demonstrate camera placement and coverage against a client's site drawing — typically requested as a deliverable ahead of, or alongside, the CCTV system quotation. It's desk/software work, not physical installation, so only Engineering hours carry the variable rate; PM, FAT, Assembly, Training, and Test & Commissioning are all zero for this row.
| Basis | Eng hrs (per qty) | One-time overhead hrs | One-time overhead cost |
|---|---|---|---|
| 1 x camera | 1 | 12 | 0 |
Quantity = number of cameras in the study.
Eng hrs (1/camera) — the actual JVSG camera placement and coverage calculation, once the site model is already set up. Scales linearly with camera count.
One-time overhead (12 hrs, flat) — work that happens once per study regardless of camera count, added once no matter what quantity is entered:
- Importing and scaling the client's site floor plan(s) in JVSG, and setting up the camera/lens type library for the models being proposed (≈ 8 hrs)
- Compiling the final coverage report and QA review before issue (≈ 4 hrs)
One-time overhead cost is left at 0. JVSG software licensing (≈ USD 600/yr) is an annual fee, not a per-study cost, so it's tracked separately as its own commercial line item rather than folded into this labor estimate.
Worked example — 10 cameras¶
| Component | Qty | Hours |
|---|---|---|
| Eng (per camera) | 10 | 10 |
| One-time overhead | 1 (flat) | 12 |
| Total | 22 |
22 total hours ≈ 2.2 man-days at the estimator's default 10 working-hours/day, ≈ SGD 1,100 in internal labor cost at the current SGD 50/hr man-hour rate.
Revision history¶
- One row: 1 hr/camera + 12 hrs flat overhead, basis
1 x camera(this page) — supersedes an earlier two-row model (a separate "Setup & Report" row summed with a "Per Camera" row) once hote added a dedicated one-time overhead field to each factor row, making a second row unnecessary. An intermediate version used1 x eafor the basis instead of1 x camera, on the assumption the unit field had to come from a fixed list — the Manage Factors unit field is actually editable, so1 x camerais clearer and is what's live now. - Both versions supersede an earlier, informal rule of thumb of 2 man-hours/camera used in ad-hoc BOQ pricing, which didn't separate fixed setup effort from per-camera effort and so over-estimated large studies and under-estimated small ones.
PAGA Sound Coverage Study (CadnaA)¶
A PAGA sound coverage study is the CadnaA acoustic modelling exercise
used to verify public address/general alarm speaker coverage and
audibility against ambient noise levels across a vessel or facility —
typically requested as a deliverable ahead of, or alongside, the PAGA
system quotation. There is no historical reference for this specific
study in the old 04 Engineering Cost Estimator R3.4.xlsm workbook —
it only has factors for designing/installing the PAGA system itself
(cabinet, speaker network wiring), not the acoustic study. There's also
no published industry benchmark for man-hours per speaker in CadnaA —
this is treated the same as the CCTV/JVSG figure: an internal estimate
based on how the team actually works, not a looked-up number.
The real CadnaA workflow has two distinct kinds of effort:
- Ambient noise field survey — walking the vessel/facility with a sound level meter, taking grid-spaced measurements (denser in noisier areas). This scales with area/zone complexity, not speaker count.
- Per-speaker modelling — each speaker's position, power, octave band, height, tilt, and azimuth is entered into the CadnaA model and verified against the propagation simulation. This genuinely is per speaker.
Same shape as CCTV, so it uses the same one-time-overhead mechanism:
| Basis | Eng hrs (per qty) | One-time overhead hrs | One-time overhead cost |
|---|---|---|---|
| 1 x speaker | 0.5 | 12 | 0 |
Quantity = number of speakers in the study.
Eng hrs (0.5/speaker) — entering and verifying one speaker's data in CadnaA once the site model already exists.
One-time overhead (12 hrs, flat) — the ambient noise field survey, site model setup, and final coverage report/heat-map compilation, added once no matter how many speakers are in the study. Reused the same 12-hour figure as CCTV on the assumption the setup/report effort is comparable in scale — revisit this once we have a few real studies to compare against, since PAGA's field survey is genuine on-site fieldwork (not desk-only like JVSG's floor-plan import), so it may turn out to need more than 12 hours.
One-time overhead cost is left at 0 — no separate software licensing or third-party cost identified for this study.
Worked example — 30 speakers¶
| Component | Qty | Hours |
|---|---|---|
| Eng (per speaker) | 30 | 15 |
| One-time overhead | 1 (flat) | 12 |
| Total | 27 |
27 total hours ≈ 2.7 man-days at the estimator's default 10 working-hours/day, ≈ SGD 1,350 in internal labor cost at the current SGD 50/hr man-hour rate.
UHF Link Budget Study (Excel)¶
A UHF link budget study verifies that each remote antenna's radio link back to the repeater/base closes with adequate margin — transmit power plus antenna gains minus feedline/path losses, checked against the receiver's sensitivity threshold. We do this in Excel rather than a dedicated RF planning tool like iBwave, which matters for the estimate: there's no 3D site model, terrain/building propagation simulation, or floor-plan import step — once the distance and path data for a link are known, it's a formula calculation in a template, not a simulation run. That makes it lighter work per unit than either CCTV/JVSG or PAGA/CadnaA.
Same situation as the other two studies: no reference in the old
04 Engineering Cost Estimator R3.4.xlsm workbook (it only has
brackets for designing the UHF repeater network itself, by antenna
count range — not a link budget study), and no published industry
benchmark for man-hours per link. hote had no UHF factor rows at all
before this one.
| Basis | Eng hrs (per qty) | One-time overhead hrs | One-time overhead cost |
|---|---|---|---|
| 1 x antenna | 0.5 | 6 | 0 |
Quantity = number of antennas (each = one link back to the repeater/base).
Eng hrs (0.5/antenna) — gathering the distance/path data for one link, entering it into the spreadsheet, and verifying the margin against the receiver sensitivity threshold.
One-time overhead (6 hrs, flat) — gathering overall site layout/antenna distances, setting up the Excel template, and writing up the final report, added once regardless of antenna count. Set lower than CCTV/PAGA's 12 hours since there's no software modeling setup or field survey step — just desk-based formula work.
One-time overhead cost is left at 0 — no separate software or third-party cost identified for this study (unlike JVSG/CadnaA, this one doesn't even carry a software licence to consider).
Worked example — 15 antennas¶
| Component | Qty | Hours |
|---|---|---|
| Eng (per antenna) | 15 | 7.5 |
| One-time overhead | 1 (flat) | 6 |
| Total | 13.5 |
13.5 total hours ≈ 1.35 man-days at the estimator's default 10 working-hours/day, ≈ SGD 675 in internal labor cost at the current SGD 50/hr man-hour rate.
Cabinet 42U Assembly — standardized to 5 man-days¶
Every system that has a "Cabinet 42U" factor row (ACS, CCTV, IDS, LAN,
NAVAID, PABX, PAGA, WMS) previously had its own assembly_hours value,
ranging from 10 hours (IDS) to 40 hours (LAN) — set independently over
time with no consistent basis. All eight are now standardized to
50 hours (5 man-days).
Why standardize instead of keeping per-system values: the actual spread across the systems people normally compare (LAN 40 / CCTV 30 / PABX 30 / PAGA 30 hrs) was modest — not different enough to justify the ongoing overhead of maintaining eight separate numbers. A market check before deciding: no published source gives exact "hours to assemble a PAGA/CCTV/LAN/PABX cabinet" (this is contractor know-how, not public data), but the NECA Manual of Labor Units — the standard electrical-contracting labor reference — puts panel wire termination at roughly 5 minutes per wire including dressing, tagging, and terminating; a 42-pole panel with ~91 wires works out to ~7-8 hours for terminations alone. That's a useful way to reason about relative complexity even without exact figures: PAGA cabinets are genuinely the most termination-heavy of the four in principle (a marine PAGA cabinet typically has one supervised loop per zone/deck — often 10-30+ zones, each with its own terminal block, zone-relay wiring, and EOL-resistor supervision loop), while LAN/CCTV cabinets have comparatively few discrete field terminations (many connections are pre-made patch cables). Despite that, we chose one flat number for all systems rather than giving PAGA a higher default — simpler to manage, and a specific project where PAGA (or any system) genuinely needs more can still be bumped up via the Calculate table's per-row override without touching the shared standard.
Why 50 hours specifically: this matches a number that already
existed elsewhere in hote — the catalog service CA42 "Cabinet
Assembly 42U" (used at the BOQ/configuration level, linked to cabinet
products like PRD-000021) was already 50 hours/5 man-days. Standardizing
the Engineering Estimator's per-system figure to the same number means
the same real-world activity (cabinet assembly) is now estimated
consistently whether you're doing a top-down Engineering Estimator pass
or pricing an actual cabinet line item in a proposal's configuration.
Subcontracted assembly costing more doesn't need a separate mechanism: if a specific project's cabinet assembly is subcontracted out at a higher cost than in-house labor, that's handled per-project — via the Calculate table's row override (Engineering Estimator) or the existing manual cost override on the configuration line item (BOQ level) — without changing this shared standard default.
Wi-Fi Coverage Study (Ekahau)¶
A Wi-Fi coverage study is the Ekahau predictive design exercise used to determine access point placement and count needed to meet coverage/SNR targets across a floor plan — typically requested as a deliverable ahead of, or alongside, the WLAN system quotation. Workflow: import or draw the floor plan, define wall materials, build the AP model library, then iteratively place APs and re-check the heatmap until targets are met, producing heatmaps, an AP bill of materials, a channel plan, and a coverage report.
Two things worth noting about this one specifically:
- hote had two separate systems for this —
WIFI("Wireless Fedility" — a typo) andWLAN("Wireless Local Area Network") — both actively used in real proposals, neither with any factor rows. Rather than resolve the duplication now, the same factor was added to both so neither is left without one; cleaning up the duplicate lookup entry is a separate task. - Unlike JVSG/CadnaA, where camera/speaker count is usually decided upfront and each unit is placed once, Ekahau's predictive design is iterative — AP count is often an output of hitting coverage targets, not a fixed input, and placement gets adjusted and re-checked several times. The per-AP figure below absorbs that iteration; it isn't as strictly linear as JVSG's one-pass-per-camera placement.
No reference exists in the old 04 Engineering Cost Estimator R3.4.xlsm
workbook (no Wi-Fi/WLAN/Ekahau content at all), and no published
industry benchmark gives hours-per-AP — market rates found were
project-level ($1,000–$5,000+ per survey, ~$2,000/10,000 sq ft), not
per-unit labor figures.
| Basis | Eng hrs (per qty) | One-time overhead hrs | One-time overhead cost |
|---|---|---|---|
| 1 x ap | 0.5 | 8 | 0 |
Quantity = number of access points.
Eng hrs (0.5/AP) — placing and configuring one AP in the model (position, height, mounting, antenna pattern, channel plan) and iterating against coverage targets.
One-time overhead (8 hrs, flat) — floor plan import/calibration, wall material and AP model library setup, and the final heatmap/ coverage report, added once regardless of AP count. Set between UHF's 6 hours (pure spreadsheet, no modeling) and CCTV/PAGA's 12 hours (which include a physical field survey) — Ekahau's setup is more involved than a spreadsheet, but this is a predictive-only study with no on-site survey step.
One-time overhead cost is left at 0 — no separate software or third-party cost identified for this study.
Worked example — 20 access points¶
| Component | Qty | Hours |
|---|---|---|
| Eng (per AP) | 20 | 10 |
| One-time overhead | 1 (flat) | 8 |
| Total | 18 |
18 total hours ≈ 1.8 man-days at the estimator's default 10 working-hours/day, ≈ SGD 900 in internal labor cost at the current SGD 50/hr man-hour rate.
RadHaz Study — flat, not per-source¶
A RadHaz (radiation hazard) study for an FPSO determines the clearance/ exclusion zones needed around RF-emitting equipment — Radar, VSAT, GMDSS, and similar — so personnel aren't exposed to unsafe RF fields. This one is modeled differently from the other four studies on this page: flat, not per-unit.
Why flat instead of per-source: the underlying calculation genuinely is per-source (each transmitting antenna gets its own safe-distance figure from output power, cable loss, and antenna gain, per RF safety literature), then combined across co-located sources to find the overall hazard zones at each deck location. But unlike CCTV/PAGA/UHF/ Wi-Fi — where the count driving the study (cameras, speakers, antennas, APs) varies widely by vessel size, often by an order of magnitude — a typical FPSO's RF-source inventory for RadHaz is small and fairly consistent across projects of a similar class (a couple of radars, VSAT, maybe GMDSS/HF — not "dozens"). There's no natural count a user would type in at estimate time the way they would for cameras or speakers, so a flat number fits better than forcing a per-unit structure onto something that doesn't really vary.
Where it lives: RadHaz doesn't belong to any single existing system in hote — it spans radar-type systems (CARS/DOPRAD/EWRS/REWS), VSAT, GMDSS, etc. all at once. It's homed under ESE ("Engineering Studies And Explorations") instead, a system that already existed with zero factor rows and reads as built for exactly this kind of cross-cutting, non-device-specific study — matching the precedent of hote's existing "CABINET" system (Cabinet & Assembly), which is also not tied to one specific piece of equipment.
How "flat" is enforced mechanically: the entire figure sits in
one_time_overhead_hours, not eng_hours — that field is designed to
never be multiplied by quantity, so the number can't accidentally
scale even if someone sets qty to something other than 1. Basis is
1 x study.
| Basis | Eng hrs (per qty) | One-time overhead hrs | One-time overhead cost |
|---|---|---|---|
| 1 x study | 0 | 16 | 0 |
16 hours ≈ 1.6 man-days at the estimator's default 10 working-hours/day, ≈ SGD 800 in internal labor cost at the current SGD 50/hr man-hour rate — regardless of how many RF sources are actually on board. A project with an unusually large or small RF inventory can still have this bumped via the Calculate table's row override without changing the shared standard.