Skip to main content

Trainer's guide outline

This is the instructor-facing companion to the Training overview. It frames the curriculum's pedagogical flow and provides a one-line summary of every module and topic so an instructor can plan a session, pace a class, or drop into a single lesson out of context.

Overarching flow​

The curriculum builds in three arcs.

Arc 1 — Conceptual foundation (Module 1). Use the calculator at calculator.siter.app to teach how the engine consumes inputs and produces a result. The calculator is a single-analysis UI: no geometry, no project, no GIS noise. Trainees see input → output cleanly and build a mental model that carries forward into Siter. This is also the natural place to teach the path triad (criteria / analysis / calculation) and the engine's nodular criteria-flow model — both surface most cleanly in a single-pair view.

Arc 2 — Working in Siter (Modules 2–7). Each module adds one layer of GIS context: access → projects → layers → features → analysis → relationships. Trainees build, run, diagnose, and mitigate, in that order. Module 4's "Adding layers" page now subsumes both the flavor decision and the add/remove mechanics — spend extra time there, since that's where the silent-omission failure mode lives. Module 7 is where most "first violations" come from being unrelated rather than from real conflicts — frame this honestly so trainees do not over-correct geometry that did not need correcting.

Arc 3 — Production and scale (Modules 8–13). Deliverables, lifecycle, utilities, bulk tools, placement preview, and external data. Module 10 and Module 11 in particular benefit from being taught after trainees have felt the pain bulk tools solve.

Recurring themes to reinforce​

Return to these whenever a trainee has a "why did the engine do that" question:

  • The triad of paths. Criteria path = the criteria references the engine cites. Analysis path = the human-readable narrated explanation. Calculation path (or logic flow) = the engine's general procedural steps. Keep these distinct in your language; trainees will mirror you.
  • Type code first. Every analysis decision starts with the type code. Whenever a result surprises a trainee, ask "what's the type code?" before anything else.
  • Same-line vs parallel under DCMA. The single most common modeling mistake. Hit it explicitly in Module 1's barricading exercise and revisit in Module 7.
  • Result sets exist on purpose. Drawings reference snapshots, not live state. Trainees who skip this end up with deliverables that mutate underneath them.
  • Criteria are nodular, not tabular. The engine walks a graph of criterion entries where each entry directs to the next based on inputs. Flowcharts visualize the actual structure, not a derived view of an underlying table.
  • Barricading is a flag, not a feature. Siter has no barricade feature type. Users assert "this is barricaded" via the segment-level flag (Segments tab) or the per-pair flag (Relationships tab). The criteria still gate the reduction. See Barricading in Siter.
  • All Siter layers are vector. The built-in base map is the only raster surface. "Custom imagery" use cases substitute vector reference content (DWG polylines, shapefile parcels) — see Background imagery.

Module 1 — Calculator fundamentals​

Pure-conceptual module. Use the calculator to teach the engine's calculation flow without the distraction of a project or map.

  • Module overview — frames the calculator as a single-analysis sandbox and previews the topics.
  • Type codes and feature categories — the first thing the engine resolves; gates which attributes apply and which analysis codes are reachable. The two terms are interchangeable.
  • Segments and sides — sides are named faces (calculator + GIS); segments are angular slices on the map (GIS only). Multi-face features without a front fail to analyze.
  • Attributes — scoped to the type code (not to PES vs ES). Cover Heavy Wall / Heavy Roof and Combustible. Introduce Most Likely vs Most Conservative defaults.
  • Relationships — gating condition for several reductions, including barricaded ILD. Operating-line context starts here.
  • Barricading — when a barricade reduces required distance and when it does nothing.
  • Explosives — hazard-division entry; some divisions take supporting weights (1.2.1 → MCE; 1.2.3 → MCE + Parenthetical Distance) alongside NEW.
  • QD results: required, blast, frag — reading the results panel; Required is the largest of the hazard-specific values. Each result returns its own analysis path and criteria path.
  • DQ analysis (actual distance) — inverts QD: given actual distance, returns allowable NEW. Capacity story alongside the QD compliance story.
  • Criteria path — the criteria references the engine cites. Distinct from analysis path and calculation path.
  • Tracing analysis paths — the human-readable narrated walkthrough. The diagnostic tool of choice for surprising results.
  • Flowcharts — per hazard division (1.1, etc.); the engine's actual decision graph rendered, not a derived view of a table.
  • Analysis codes — the facility-side classification (AGSU, ECM7FB, IBD, ILDU) — the engine's answer to "how do I treat this facility?" Each side of a pair gets its own analysis code, which then feeds the HD libraries that produce the actual K-factors and required distances.
  • Exercise: Barricaded related vs barricaded sameline — demonstrates K9 vs K18 under DCMA: a barricade reduces ILD only across the same operating line, not across parallel lines.

Module 2 — Getting set up​

Quick access-plumbing module. Get trainees past the login wall and editing.

Module 3 — Projects​

The first Siter-native lesson. Frame project-creation choices as durable decisions — several are hard to change later.

  • Creating a project — coordinates, criteria, default attributes, spatial reference. Spend real time on Most Likely vs Most Conservative — it shapes the whole project's behavior.
  • Importing a project — bringing in a snapshot. Note that user permissions are not preserved on import.

Module 4 — Layers​

Foundational organizing concept. The QD-visible vs background distinction is a frequent source of "why isn't this analyzing" confusion later — establish it firmly here.

  • Adding layers — flavor decision (QD-visible vs background), add/remove mechanics, silent-omission failure mode.
  • Organizing layers — order, visibility, naming conventions that scale.
  • Exporting layers — handing off a subset of project data.
  • Importing layers — GeoJSON (UTM or WGS 84) and shapefile (any proj4 CRS, zip with sidecars). User picks flavor, default type code, and name column at the import dialog.
  • Background imagery — built-in base maps; custom imagery use cases go through vector reference layers (no raster layer support). Defer the vectorization workflow to Module 13.

Module 5 — Features​

Single-feature CRUD. Bulk versions live in Module 11; resist teaching them here.

  • Adding features — the four shape-creation methods (Digitize, Place Template, Known Position, Reshape) and type-code-at-creation.
  • Editing features — the dashboard tabs and which edits invalidate prior analysis.
  • Moving features — geometry edits and layer reassignment, including the silent-omission symptom of a wrong-flavor layer.

Module 6 — Your first analysis​

Pivot module. After this, trainees have run and read a real analysis.

  • Project analysis results — the project results panel (filter views, columns, per-row detail), the per-feature Calculations tab (filtered to that PES), and the Spatial Analysis tab (encumbrances).
  • Analysis symbols — four symbol categories (arcs / spatial-analysis encumbrance lines / violation lines / dimension lines), per-result-set styling, the union toggle on arc styles.
  • Violations — distinguishing real conflicts from artifacts of missing data; includes the Requires Analysis filter as a diagnostic companion (a non-empty list right after a run is the signal worth investigating).
  • Exercise: Violations on what should be related facilities — reinforces "many violations are data-completeness issues, not siting problems."

Module 7 — Relationships​

The single biggest lever for resolving violations. Set expectations: most trainees will mitigate more violations here than they will by moving geometry.

  • Individual facility editing — the per-feature dialog has two halves (related-features picker + group-membership chips); codes come from the matrix, not from this dialog.
  • Relationship groups — the per-criteria matrix authors the codes; the Build action materializes the matrix into pairwise relationships across all encumbrances. Spatial analysis must run first.
  • Barricading in Siter — a flag, not a feature. Set per-segment (Segments tab) or per-pair (Relationships tab). Same K9 vs K18 calculator-side rules still gate the reduction.
  • Exercise: Mitigating violations via relationships — full-project walkthrough using only relationship and barricading edits.

Module 8 — Making site plans​

Production deliverables. Drive home the result-set-first workflow.

  • Forms — per-feature, per-criteria submittal documents. Three sections (General Information / Data on Facility / PES-ES Paired Relationships), dual-direction reporting (forward + Reverse), inline criteria citations as the audit trail. Forms are ephemeral — regenerated each view, so freeze them against a result set for external delivery.
  • Result sets — frozen analysis snapshots with their own four-tab structure (Details, Calculations, Spatial Analysis, Symbols) and per-result-set styling. Used for stable deliverables and for the COA (course-of-action) comparison workflow — capture one result set per scenario.
  • Drawings — HTML pages opened in a new browser tab. Four scale presets (1":400' / Current, each with optional Legend); inherit per-result-set symbology when generated from a result set.

Module 9 — Project management​

Lifecycle and collaboration. Quick module — mostly mechanics.

  • Sharing — token-based flow (generate token → recipient submits request → admin accepts with a chosen role). Four project roles: Read Only / Read and Write / Just Administrator / Full Administrator. Online only; desktop is single-user.
  • Export and import — round-trip with the permissions-not-preserved and undo-stack-not-preserved caveats.
  • Copy — same-environment duplicate for exploring alternatives or seeding a COA.
  • Archive and delete — archive removes from the active list but keeps the data; delete completely destroys the project.

Module 10 — Utility tools​

Quality-of-life tools. Best taught after trainees have run a project for long enough to want them.

  • Data validation — Explosive Scrubber removes orphaned explosive entries.
  • To-do list — in-project task tracking.
  • Markup — distance and buffer annotations on the map.
  • Digitizer — precision drawing with snapping and constraints; GPS traces are planned but not yet available.

Module 11 — Bulk tools​

Scale tools. Trainees should feel the pain of one-by-one editing before this module — that motivates the lesson.

Module 12 — Placement analysis​

Quick module. The "what-if" workflow before committing a feature.

  • New feature preview — drag a prospective feature and watch arcs and violations update live.

Module 13 — Advanced data​

External data ingest. Treat as reference material — most trainees will only use one or two of these.

  • Shapefile import — .zip packaging requirements with full sidecar set.
  • GeoJSON import and export — the preferred interchange format for ad-hoc moves; UTM or WGS 84 only.
  • AutoCAD — .dwg / .dxf interchange for projects whose geometry lives in CAD.
  • QGIS — vector prep for Siter imports (CRS conversion, format conversion, vectorizing scanned drawings). Siter does not import rasters as layers, so QGIS's role is always upstream of a vector export.
  • ESS imports — bringing an ESS export into Siter as a starting project.