Semarang is not a greenfield. It already runs BRT, ATCS, flood sensors, and a decade of digital services — fragmented across fifty apps while the coast subsides. This master plan integrates what exists, puts resilience first, and treats operations as the product.
The systems exist. What's missing is integration, operations, and the right priority order. Phase 0 money goes into plumbing — not new apps.
Most Indonesian smart-city programs fail at procurement and operations, not technology. These rules are the difference between a platform and a graveyard of dashboards.
Subsidence and tidal flooding are existential. Hydromet sensing, pump SCADA, and InSAR monitoring are Tier-1 — chatbots are not.
Fifty apps already exist as islands. One data platform, one identity, mandatory APIs — before any new point solution is bought.
Every capex item carries a funded 5-year O&M plan, city-owned data and code, and an exit clause. No lock-in, no dead dashboards.
Every domain rides the same plumbing. The program map below is the whole plan on one screen — explore each domain in detail on the next page.
The plan is gated, not continuous. Each phase ends at a go/no-go checkpoint, and Gate 1 is timed before the 2030 mayoral transition on purpose — political risk is the single biggest threat to a decade-long program.
Each phase carries its own deliverables and exit condition. Phase 0 has a kill switch; the others have outcome targets.
Phase bands, the three gates, and the mayoral transition on a single axis.
Public transport, traffic, waste, flood & environment, public services, and asset management — each riding the same shared data platform, identity, and sensing network underneath.
Each domain leads with a measurable goal, then the programs that deliver it, then the KPIs it reports.
A plan is only as good as its money, its metrics, and its honesty about what could break it. Here are all three.
About Rp 280–350 B per year, roughly 4–5% of the city's APBD. Self-financing programs — streetlight ESCO, smart parking, property-tax intelligence — are front-loaded so the program pays for its own later phases.
Baseline → 2035. Teal targets rise; terracotta targets fall. The public scorecard is the enforcement mechanism — not goodwill.
Named openly, because the mitigations are structural — not hopeful.
Embed in Perda + RPJMD; Gate 1 timed before the election; visible citizen wins early.
Perda mandate, Sekda-level escalation, OPD KPIs tied to performance pay.
City owns code + data; 20% O&M budgeted; exit plan in every contract.
Cheap replaceable LoRaWAN units; local SME maintenance; EAM asset tags.
DPO + privacy impact assessments, ISO 27001, pen-testing, CSIRT.
Sensing tied to binding groundwater & building permits — dashboards alone change nothing.