If you’re managing a fibre deployment — or considering outsourcing your design work — you need to understand the difference between High-Level Design (HLD) and Low-Level Design (LLD). Not in theory, but in terms of what you’re actually getting delivered.
This isn’t academic. The handover between HLD and LLD is where most projects lose time, because the expectations on both sides are poorly defined.
What is High-Level Design (HLD)?
HLD is the strategic phase. It defines the overall network architecture before anyone puts boots on the ground. Think of it as the blueprint that answers: how will this network work, and roughly where will everything go?
A typical HLD deliverable includes:
- Network topology — ring, tree, or mesh architecture. Where the OLT sits, how many distribution points, what the splitter ratios look like.
- Coverage area definition — which premises are in scope, grouped by exchange area or distribution zone.
- Route planning — high-level duct routes based on desktop survey. Major road crossings, trunk routes, existing infrastructure that can be reused (PIA, existing duct).
- Splitter placement strategy — 1:32, 1:64, centralised vs distributed splitting. This decision has massive cost implications downstream.
- Bill of Materials estimate — preliminary BOM based on design assumptions. Not construction-ready, but enough for budgeting.
- UPRN counts — premises passed per zone, with take-rate assumptions applied.
HLD is typically done from the desk using GIS data, OS mapping, Openreach PIA records (in the UK), or equivalent data sources. Some providers supplement with Google Street View.
The output is a set of documents and GIS layers. Not a construction pack.
What is Low-Level Design (LLD)?
LLD takes the HLD and turns it into something a construction crew can actually build from. This is where survey data meets engineering detail.
A typical LLD deliverable includes:
- Detailed route design — exact duct routes, chamber locations, pole positions. Every splice point, every joint, every customer drop.
- Survey validation — LLD requires physical walkout or drone survey. Desktop assumptions from HLD get confirmed or corrected.
- Splicing schedules — fibre allocation sheets showing exactly which fibre goes where, from OLT port to customer premises.
- Construction packs — the documents a civils contractor needs: dig plans, reinstatement specs, traffic management requirements, permit applications.
- As-built preparation — design structured so that as-built records can be captured during construction.
- Final BOM — exact quantities of cable, connectors, chambers, cabinets. This is what procurement orders against.
LLD is significantly more labour-intensive than HLD. A single city that takes 2-3 weeks at HLD might take 2-3 months at LLD.
Where projects go wrong
1. Treating HLD as “good enough” for construction
HLD gives you 70-80% of the picture. Some operators try to skip LLD to save time. The result is field crews making design decisions on the fly, inconsistent build quality, and as-built records that don’t match reality.
2. No clear handover specification
If your HLD team and LLD team are different groups (common when offshoring), you need a precise specification for what gets handed over. File formats, coordinate systems, naming conventions, attribute data in GIS layers. A vague handover kills velocity.
3. Survey scope creep in LLD
LLD walkouts can reveal that HLD assumptions were wrong — a duct route is blocked, a pole can’t take additional loading, a road crossing needs a different approach. Good LLD teams flag these early. Bad ones redesign silently and blow the budget.
How this affects offshoring
HLD is highly suited to offshore delivery. It’s desk-based, data-driven, and the quality can be verified against source data. Many Indian design teams produce excellent HLD work at a fraction of the onshore cost.
LLD is harder to offshore completely because of the survey component. The most effective model is:
- Onshore: Survey teams doing walkouts and data capture
- Offshore: Design teams producing detailed construction packs from survey data
The critical handover point is the survey data package. If it’s well-structured and complete, offshore LLD works brilliantly. If the survey team captures incomplete data, you get costly rework cycles.
The bottom line
HLD and LLD are not two versions of the same thing — they’re fundamentally different phases with different skills, different timelines, and different deliverables. Getting the boundary clear between them, and the handover specification tight, is one of the highest-leverage things you can do for project delivery.