LuxFTTH Design Software: A Buyer's Guide for Fiber Teams
Written by
Lux Team
Category
Guides
Date
Share this article
The four kinds of tools sold as FTTH design software
Search for FTTH design software and you get list articles that rank the same products in a different order every time. That is because they compare tools that do different jobs. Before you compare vendors, decide which job you are buying for. Most fiber teams end up using tools from three or four of these categories.
CAD. AutoCAD, MicroStation and similar tools draw construction prints well, and many permit offices still want DXF or DWG. A CAD drawing knows nothing about the network, though: a line is a line, not a 144-fiber cable with a splice at each end.
Desktop and enterprise GIS. QGIS, ArcGIS Pro and utility network models keep geometry and attributes in one place and handle map production well. Fiber connectivity, meaning which strand lands on which port, is usually bolted on or kept in spreadsheets.
Fiber planning and optimization engines. These take demand points, a street graph and design rules, then propose splitter locations and routes. They are strong at high-level design and at comparing scenarios. They vary a lot in how far they carry the design toward construction.
Fiber management and records systems. These are the record of what is built, down to the strand. They are often where the as-built ends up and where operations looks for a fiber path during an outage.
The expensive mistake is buying one category and assuming it covers the others. A planner that produces a beautiful high-level design still leaves someone to draw splice sheets, size the bill of materials and redraw everything when the field survey moves a pole. If you are new to the design stages themselves, read our guide to HLD vs LLD in FTTH design first.
Also ask how each tool is priced: per seat, per home passed, per project or per design. Per-seat pricing punishes you for giving field and permit staff access; per-home pricing can punish large overbuilds that you may never construct. Model the cost for your next two years of work, not for the pilot.
Seven questions that separate the tools
Feature checklists all look the same. These questions expose real differences in a demo.
What happens to everything downstream when one thing changes? Move a cabinet 50 m in the demo. Do the routes, cable lengths, splice sheets, bill of quantities and permit drawings update, or does someone redraw them? Change, not first-draft speed, is where design hours go.
Where do your rules live? Drop length limits, homes per closure, splitter cascade depth and optical budget should be rules the tool checks, not a PDF your engineers remember. Ask to see a violation flagged and explained.
Does the loss budget use routed lengths? A budget calculated on straight-line distance understates the real path, because on a street grid cable follows the streets and the poles, not the crow. Ask how the tool measures the worst path from OLT to home.
Is pricing yours? A bill of materials priced from a vendor's generic catalogue is a guess. You want your own rate card and unit costs, with project-level overrides.
What leaves the tool, and in what format? At minimum, look for DXF for permit and CAD teams, shapefile or GeoJSON for GIS, KML for quick review, and Excel or CSV for the bill of quantities. If the only way out is a proprietary export, you are buying lock-in.
Who approves, and is it recorded? Someone with an engineering licence signs the design. The tool should record who approved which version, and a change after approval should create a new version, not silently overwrite the old one.
Is it built for one domain or your standards? Colour codes (TIA-598 in North America, other schemes elsewhere), closure types, fiber types such as ITU-T G.652.D in the feeder and bend-insensitive G.657.A2 in drops, and labelling conventions differ by operator. Good software lets you change these without a vendor ticket.
Run a pilot, not a demo
Demos run on data the vendor has tuned for months. Before you sign, run a two-week pilot on one of your own serving areas, ideally 500 to 1,000 homes with a mix of aerial and underground plant.
Give every vendor the same inputs:
address points or building footprints for the area;
a pole layer and existing duct, even if it is incomplete;
your design rules on one page: split ratio and architecture, maximum drop length, homes per terminal, spare fiber policy and optical budget class (for example GPON Class B+, 28 dB);
your unit-cost rate card.
Then score what comes back:
Is every home connected, or are there orphans?
Does the design break any of your rules? Check five closures by hand.
Does the bill of materials reconcile with the design? Count cabinets and splitters yourself.
How long does a realistic change take: moving a cabinet, swapping a pole route for underground, adding a 24-unit building?
Can your GIS, CAD and permit people open the exports without help?
Write the scores down before the pricing conversation. Teams that skip this step usually buy on interface polish and find the gaps during the first construction season.
Red flags during a pilot:
the vendor asks to clean or reshape your data before the run, and will not show you what changed;
rule violations appear only in a separate report, not on the design itself;
the bill of materials is a template with quantities filled in, rather than counted from the design;
an export opens in your GIS but loses attributes such as fiber counts or cable IDs;
approvals are a status field anyone can edit, with no history of who changed what.
Where AI agents fit, and what Lux does today
The newest category is agents: software that does the drafting work an engineer would do, then stops for review. This is the approach Lux takes. Lux is built for the non-physical work of a fiber network, from design and splice documentation to permits and records. Engineers approve everything.
In a Lux project, the agent proposes a plan first: the architecture template, the order of work and the expected counts. A person accepts it before any drafting starts. The agent then:
places closures with an optimizer and states it plainly if it had to fall back to a simpler method;
places cabinets;
routes drops, distribution and feeder, aerial on existing poles where your rules allow;
checks the result against the project's design rules, including length limits measured along the routed path and the optical budget on every OLT-to-home path;
compiles the bill of materials, the bill of quantities and the work orders, priced from your rate card.
Move a cabinet and only its cables re-route; the bills and checks recompile. Approval is a recorded, versioned step, and it only opens when the rule check passes.
Designs export to DXF, shapefile, KML, GeoJSON and CSV, with bills and work orders in Excel and PDF. Lux is not a replacement for every tool on your desk; it is meant to do the repetitive drafting between them. See how it works on the design platform page, or the network design service if you want our team to deliver designs on it.
Run your pilot area through Lux: send us a serving area and your design rules, and we will show you the design, checks and bill of materials on your own data. Talk to the Lux team.
Frequently asked questions
What is the best FTTH design software?
There is no single best tool, because CAD, GIS, planning engines and fiber records systems do different jobs. The best choice is the one that handles your change workflow, checks your design rules and exports the formats your permit and GIS teams need. Test it on your own serving area.
Is QGIS enough for fiber network design?
QGIS handles geometry, attributes and map production well and costs nothing to license. It does not natively model strand-level connectivity, splitter cascades or optical budgets, so most teams add plugins, scripts or a fiber-specific tool alongside it.
Can FTTH design be automated?
Much of the drafting can: terminal placement, routing, rule checks, loss budgets and bills of materials follow rules a computer can apply. Engineering judgment, field verification and the final sign-off still need a qualified engineer.
What is the difference between fiber design and fiber management software?
Design software produces a network that is not built yet. Fiber management software is the record of what was built, down to which strand lands on which port, and is used by operations. Some tools cover both, but ask how a design becomes the as-built record.
Share Blog
