FiberBuild designs the optical network of a whole zone (closures, splitters, FATs, cables, fibers, splices and slack) and delivers the A1 drawings, the DWG and DXF, the KMZ, the fiber diagram, the Excel of boxes and splices and the bill of materials with costs. Then it takes that design all the way to the live network, inside your ISP CRM.
Demo zone: "Colonia Demo", a fictional neighborhood over real streets of Celaya, Guanajuato. Every figure comes from a file the factory generated.
The Excel with the coordinates and the profile with your conventions go in; everything the crew and the build manager need comes out. The video walks through a real demo zone from start to finish, with subtitles and no sound.
No reworking your data for the system: it accepts inputs as you already have them, and whatever changes from one company to another lives in a profile you edit, not in the code.
H1R1T1, H1R1-1D1, CTO-001…estandar: fisher_networksarquitectura: spl1: 8 # cierre T spl2: 8 # FAT fat_puertos: [8] costos: {cierre: 2500, m_distribucion: 14, m_troncal: 19}cables: troncal: {nombre: "ADSS 96H", hilos: 96} distribucion: {nombre: "Fig-8 24H", hilos: 24} colores_buffer: [AZUL, NARANJA, VERDE, CAFE, GRIS, ...]reservas: caja_paso: "2*H+2+1.5"optica: atenuacion_db_km: {1310: 0.35, 1490: 0.25} presupuesto_db: 28.0reglas: vano_max_m: 80nomenclatura: fisher_hr cierre: "H{hub}R{ruta}T{t}" fat: "H{hub}R{ruta}-{t}D{dc}"
A fragment of the profile used to design the demo zone. Costs are reference values: replace them with yours.
It is not a black box: each step applies a rule of your standard and leaves its result in the report, with the alternatives it discarded.
It snaps each route vertex to the nearest pole (within 10 m), inserts the poles the line crosses without a vertex and builds the span network. A span longer than 80 m calls for an intermediate pole. Then it builds the shortest-path tree from the entry point, preferring the trunk.
By coverage, every pole is within one span of a FAT. With homes per pole, ports equal homes times take rate and the smallest FAT that covers the demand is chosen; boxes are only placed where there are homes.
It splits the tree into first-level splitters and tries several groupings of T closures; the one with the lowest real cost wins: closures plus fiber to buy. If several splitters go down together along a long branch, it places a closure off the trunk with a single cable.
Splitters of the same closure going down the same street share a cable, one per buffer. Fibers are numbered in route order and every buffer and fiber carries its TIA-598 color. Each splice ends up on its own row.
Optical loss at each FAT against the budget, slack loops and storage with your formulas (2*H+2+1.5), tension and suspension hardware pole by pole, materials, and validation against the standard.
With the NOC coordinates it traces the shortest path along OpenStreetMap streets to the entry of the zone. Its length goes into the dB and into the cable to buy, and it is drawn on the plan and in the KMZ.
To show that conventions rule, the demo zone was designed again with a 1:4 × 1:16 standard, a 48-fiber feeder and CE-01 and CTO-001 naming. Same poles, a different network, and neither with validation errors.
| Fisher profile | 1:4 × 1:16 profile | |
|---|---|---|
| Splitters (1st × 2nd level) | 1:8 × 1:8 | 1:4 × 1:16 |
| FATs | 40 of 8 ports | 31 of 16 and 1 of 32 ports |
| Customer ports | 320 | 528 |
| PON ports | 5 | 9 |
| T closures | 3 | 3 |
| Distribution cables | 4 · 2,187 m | 9 · 2,695 m |
| Trunk fiber to buy | 1,505 m of ADSS 96H | 1,529 m of ADSS 48H |
| Distribution fiber to buy | 2,848 m of Fig-8 24H | 3,237 m of Fig-8 12H |
| Maximum loss at 1310 nm | 23.67 dB | 23.67 dB |
| Closures plus fiber | $75,967 | $81,869 |
| Names | H1R1-1D1 | CTO-001 |
They all come from the same design, so they do not contradict each other: the splice you see in the diagram is the same one in the Excel sheet and the same one the bill of materials counts. The images are real outputs of the demo zone.
1-2D18, FAT 8P NAR H14, 574m 23.3dB@1310.Each zone was designed using only its KMZ and compared against the design that outside-plant engineers had made by hand, in real zones. More than 50 automated tests run every time a version is packaged.
| Test | What it checks | Result |
|---|---|---|
| Regression | Reproduce box by box the design of a real zone, made with its own scripts | 62 FATs, fibers, cables, slack and hardware identical |
| Calibration | Real zone, using only its KMZ, against the DWG made by hand | 38 FATs by hand; 100 % have a factory FAT within one span or less |
| Validation (not used for calibration) | Another real zone, with the same profile and no readjustment | 41 FATs against 45 (91 %; uncalibrated it was 32, 71 %); 97 % of the hand-placed FATs covered within one span; 18 on the exact same pole |
| Closures by cost | Automatic closures against the ones placed by hand | 3 closures for $206,317 against 4 for $209,355, with the same maximum loss |
| Different inputs | Excel in degrees, minutes and seconds against KMZ | Identical design |
| Other standards | 1:4 × 1:16, 48 fibers, other naming schemes and templates | 0 validation errors |
| Fiber diagram | Fiber-by-fiber continuity | Every FAT is reached from its PON through a continuous chain of fibers and splices; no fiber repeats in a cable; every splitter output goes to a FAT or stays as spare; the splices in the materials, the model and the drawing match |
| Traced feeder | NOC to the first closure, along OpenStreetMap streets and by another algorithm | 4,953 m against 4,946 m: 0.1 % difference |
| AutoCAD AUDIT | The generated DWG, opened and audited | 0 errors introduced by the design |
The comparison zones are real and their names are not published. In one of them the person placed all three closures off the trunk and the factory does not reproduce that layout: there it proposes a different scheme, not the same one with other numbers.
A design that is delivered and forgotten stops being true the day the crew moves a box. Inside Fisher Platform the design stays alive until what is built and what is drawn match.
The ERP keeps the queue and the workers ask it for jobs: an always-on Linux server and, if you want, a laptop with AutoCAD. No ports are opened and each worker has its own token.
Every change is a revision. Compare them on FATs, PON, ports, closures, fiber, loss, cost and cost per port, and pick the good one for each project.
Each material is matched to an item in your warehouse and takes its purchase price, its average cost or its price. With no price it says "no price", never zero. With one click, a purchase request or a draft warehouse issue.
Pole inventory with owner and status (proposed, requested, authorized, installed, withdrawn), an exportable list with hardware, rent per pole or per contract with reminders, and a linked case file.
Investment in materials and labor, rent and maintenance costs, homes passed, revenue per customer, scenarios, present value and payback month, every figure with its formula and its source.
Closures, splitters, FATs, cables and splices move to the FTTX plant map as planned, without re-entering data. It can be repeated without duplicating and undone while nothing is built.
From the phone, with GPS, the crew sees poles, FATs and closures by proximity and marks what is built, moved or nonexistent, with a photo and a note.
What is marked in the field becomes an "as-built" revision and is reconciled with the plant, with a preview, blocks if customers are connected, and undo.
The FTTX plant, the NOC, the ONTs and the outages live in the same system. What you build with FiberBuild is what you later see on the plant map and what a cut affects.
The bill of materials becomes a purchase request or a warehouse issue, with your items and your prices, using Purchasing numbering.
The pole access request opens as the project's case file with the template you choose, and the pole list comes out already with the real poles.
Alerts when the design is ready, when it fails, when jobs are waiting and no worker is online, and when a pole rental contract is about to expire.
The copilot queries projects, revisions, poles and the business case, and proposes with a confirmation card: queue a design, move it to the plant, mark what is built or request the purchase.
Every finished revision, every handover to the plant and every pole status change publishes its event. Profiles, field marks and pole owners are open catalogs.
Better to know before you sign up than to find out on the build.
No. The A1 PDF drawings, the DXF, the KMZ, the Excel and the bill of materials are generated without AutoCAD. DWG output needs ODA File Converter (free) or AutoCAD on the machine that does the job.
Any that can be described in a profile: 1:8 × 1:8 cascade, 1:4 × 1:16 or another, FAT sizes, cables and TIA-598 buffer colors, slack, losses and optical budget, maximum span and hardware rules. A profile inherits from another and changes only what differs.
Yes. It starts from an Excel, a CSV, a KMZ or a DWG of poles. The utility-company drawing is optional: if you have it, the design is mounted on a copy of it, and if it is not georeferenced, the factory aligns it by comparing pole numbers.
The demo zone, with 89 poles, was designed in 5.5 seconds without drawings and in 36.9 seconds with the A1 PDF drawings, on a laptop. DWG output adds the time of the engine that converts it.
The design is validated against your standard and was tested against real zones made by hand, but it is still an engineering proposal: the person responsible for the build reviews and signs it. Every drawing carries the note to verify details on site.
The profile comes with reference prices (for example, a $2,500 closure, an $850 8-port box and fiber at $19 and $14 per meter) that you can edit. In the ERP each material is matched to an item in your warehouse and takes your purchase price, your average cost or your price.
The crew marks it from the phone. The marks become an "as-built" revision, are compared against the plant and are reconciled with a preview, blocks and undo, so the drawing and the ISP plant say the same thing as what is built.
Yes. FiberBuild is included in the Enterprise plan of the ISP CRM, at USD 120 per month, along with TR-069, inventories, WhatsApp, Telegram and AI.
FiberBuild comes included in the Enterprise plan of the ISP CRM. Send us the Excel or KMZ of one of your zones and we will return it designed, with your conventions.
All images and figures come from a fictional demo zone, "Colonia Demo", laid out over real streets of Celaya, Guanajuato, with 89 estimated poles that do not correspond to real poles. The drawings, the KMZ, the Excel and the bill of materials are real outputs of the design factory; the map is the KMZ drawn over OpenStreetMap. Prices are the profile's reference prices. No data belongs to a customer.
Costs and times depend on your standard, your prices and the machine doing the job; the 36.9-second measurement is for the demo zone on a laptop, with PDF drawings and without DWG.
Map © OpenStreetMap contributors.