Lesson 8.3Lesson 8.3 · GIS for Urban Planning
Infrastructure & Utility Planning
Water, sewer and transport as networks the GIS can trace
The new colony has water on the plan and no water in the tap. On the map, the main stops one junction short.
Everyone approved the layout; the pipe is drawn; the households exist. Yet no water arrives - because a network is not a picture of lines, it is a chain of connections, and one missing link breaks the whole downstream branch. Infrastructure is the one planning layer where connectivity, not just position, is the fact that matters - and GIS built for networks is what lets you ask the map 'does this actually reach?'
A network is only as strong as its worst-connected node.
Nodes and links: the network data model
A road system, a water grid, a sewer, a power feeder - all share one shape. They are networks: a set of nodes (junctions, valves, substations, road intersections) joined by links (pipe segments, road centrelines, cables) that carry flow. In GIS this is a specific data structure with topology - the links know which nodes they connect, and the nodes know which links meet them.
That topology is what separates a utility layer from a pretty set of lines. With it, the GIS can trace: start at the source and find everything downstream; find the break that isolated a branch; compute the shortest path or the reachable set within a distance. Without it, you have a drawing of a network, not a model of one - and a drawing cannot tell you why the tap is dry.
Two lines that touch on screen are not connected unless the topology says so.
Water and sewer as layers
Water supply and sewerage are the classic utility layers, and they behave oppositely: water is a pressurised network pushing outward from a source (reservoir, overhead tank), while sewer is largely a gravity network draining downhill to a treatment works. That difference is why sewer planning leans so heavily on the terrain layer from Module 5 - a sewer that has to run uphill needs a pump, and the DEM tells you where.
On GIS each is its own layer of links with attributes - diameter, material, capacity, laid-date - and nodes with attributes - valves, manholes, pump stations. Modelled this way, you can size a main against the demand of the plots it serves, spot where a proposed colony sits beyond the reach of an existing main, and hold an as-laid record that survives the retirement of the engineer who remembers where the pipes are.
Transport and the reach of a network
Roads are the network everyone shares, and their planning question is usually reach: what can be got to, from here, within a distance or a time? GIS answers this with network analysis - not a straight-line circle, but a trace along the actual links. The reachable set within, say, 800 m of walking is the service area (or catchment); the best route between two points is the shortest path.
This matters because straight-line distance flatters everything. A clinic 500 m away as the crow flies may be a 1,200 m walk around a rail line or a drain. Planning bus stops, deciding where a new road unlocks land, testing whether a proposed junction over-loads - these are all network questions, and getting them right needs the links, not just the points.
PM Gati Shakti: one base for every ministry
For decades each department drew its own infrastructure on its own base - the roads authority, the railways, the power utility, the water board - so lines that should meet on the ground missed on paper, and a road was dug up a month after it was laid to lay a cable. PM Gati Shakti - the National Master Plan is India's answer: a dynamic GIS platform (built on BISAG-N) that integrates more than 550 infrastructure data layers across ministries onto one shared base map.
The planning value is coordination. When every network sits on the same base, you can see the clash before it is built, sequence works so the road is laid after the pipe, and plan multimodal connectivity - port to rail to road - as one system. It is login-gated for institutional users rather than an open download, but conceptually it is the layer-overlay idea from Module 0 scaled to a nation.
One base map is worth a hundred coordination meetings.
The honest limits
Utility GIS is only as good as its record, and India's records are uneven. Much underground infrastructure was laid before any GIS existed, so 'as-laid' data is often reconstructed from memory and patchy surveys - present on the map, uncertain on the ground. Topology has to be built and checked; two lines that merely touch on screen are not connected until the data says so, and a single un-snapped node silently breaks a trace. Data still lives in departmental silos despite Gati Shakti's ambition. And demand is a moving target - a network sized for today's population is under-sized the day the FAR is raised.
So validate topology before you trust a trace, date-stamp as-laid layers, and treat capacity analysis as a scenario tied to a stated demand.
PM Gati Shakti — National Master Plan
550+ integrated infrastructure layers across ministries (BISAG-N)
India's shared GIS base for multimodal coordination; login-gated for institutional users.
OGC WFS (Web Feature Service)
Serves queryable vector features (geometry + attributes)
The interoperable way to publish and consume utility network features across agencies and tools.
ISO 19107
Spatial schema - geometry and topology model
The global standard behind proper network topology - the reason a 'connection' is well defined.
Smart Cities — DataSmart Cities
City-level open datasets, some with GIS infrastructure layers
Free open datasets from Smart Cities; the Mission closed in 2025 but the data portal remains reachable.
EPSG:32643 (WGS 84 / UTM 43N)
Projected metre-based CRS over much of India
Measure pipe lengths, buffers and service distances here - never in degrees (EPSG:4326).
Workshop — build a network and find its service area
Take a road layer, build routable topology, and compute the walkable service area around a facility - then compare it against a naive straight-line buffer.
QGIS 3.44 (free) with QuickOSM + Network Analysis, or ArcGIS Pro 3.7 with Network Analyst.
Goal: an 800 m walking service area around one facility, plus the difference from a 800 m circular buffer Data: OSM roads (via QuickOSM) + one facility point Time: ~45 minutes
- 1Load a road network and a facility point in a metre-based CRS (EPSG:32643). In QGIS: QuickOSM for roads. In ArcGIS Pro: add a street layer or OSM data.
- 2Build/validate topology so the network is routable. In QGIS: use the Processing 'Network Analysis' tools (they build the graph from the line layer). In ArcGIS Pro: create a Network Dataset (Network Analyst).
- 3Compute the service area (reachable set) at 800 m along the network from the facility. In QGIS: Processing ▸ Network Analysis ▸ Service area (from layer). In ArcGIS Pro: Network Analyst ▸ Service Area, break value 800.
- 4Draw a plain 800 m circular buffer around the same point for comparison. In QGIS: Vector ▸ Geoprocessing ▸ Buffer. In ArcGIS Pro: Analysis ▸ Tools ▸ Buffer.
- 5Overlay the two and compute the difference (buffer minus network area) to quantify how much the straight-line circle overstates real reach. In QGIS: Vector ▸ Geoprocessing ▸ Difference. In ArcGIS Pro: Analysis ▸ Tools ▸ Erase.
You’ll walk away with
A network-based service area, a comparison against the naive buffer, and a measured statement of how much straight-line distance flatters accessibility.
Three altitudes on the same idea
Read the band that fits you — or all three.
Infrastructure availability is a go/no-go for your site. Before promising a client anything, check whether trunk water, sewer and a road of adequate width actually reach the plot on the utility layers - not whether they exist somewhere nearby. A scheme that assumes a connection one junction too far is a scheme that stalls at occupancy.
You allocate the density; you must fund the network that serves it. Every FAR you raise adds load to water, sewer, drainage and roads. Model those as networks tied to the land-use layer so capacity keeps pace with permitted development, and use Gati Shakti-style shared bases to sequence and coordinate works instead of trenching the same road twice.
The road network is the skeleton of the public realm. Junction density, block permeability and walkable catchments are network properties, and they decide whether a place feels connected or severed. Use service-area analysis to test whether people can actually reach the destinations you are designing streets toward - along the links, not across them.
“If two utility lines meet on the map, they are connected and flow will pass between them.”
Do it yourself
No software needed - just think in nodes and links.
- 1Explain, in one sentence, why a water network and a sewer network have opposite relationships to the terrain layer.
- 2A clinic is 500 m away in a straight line but a 1,300 m walk. Which distance should a service-area plan use, and why?
- 3Describe what 'topology' adds to a set of pipe lines that plain geometry does not.
- 4Name three ministries whose layers PM Gati Shakti would bring onto one base, and one clash that base could prevent.
- 5Why can an as-laid utility layer be present on the map yet uncertain on the ground?
The one line to carry out
Peer-reviewed journals & authoritative standards
- 01Miller, H.J. & Shaw, S.-L. — Geographic Information Systems for Transportation: Principles and Applications — Oxford University Press, 2001.
- 02de Smith, M.J., Goodchild, M.F. & Longley, P.A. — Geospatial Analysis: A Comprehensive Guide, 7th ed. — Winchelsea Press, 2025.
- 03Batty, M. — The New Science of Cities — MIT Press, 2013.
- 04Computers, Environment and Urban Systems — Elsevier, ongoing.
- 05Habitat International — Elsevier, ongoing.
A network only earns its cost if it serves people - so the last lesson turns to who those people are and where the gaps are: population density, accessibility and gap analysis.
The author
Amogh N P
Architect, interior designer, and creative polymath. Studio Matrx began in his notebooks — his vision of design made honest, useful, and open to everyone. Its Academy is written and taught in his memory, and free, forever.
More about Amogh →