Shenzhen Smart City

Original stage labels are preserved. Unverified patterns are research audit entries, not deployments.

[]

# Shenzhen — Smart City Dispatch v0.2.3 research report

**Status:** PROVISIONAL RESEARCH · prepared 2026-10-11 · **2C — provisional, conditional operational autonomy**
**Boundary:** Shenzhen municipality including urban districts under the municipal government; do not credit neighboring Dongguan, Guangzhou, or Hong Kong robotaxi or utilities.

**Scores:** Type 24.79 (exact 24.79166666666666666666666666); Autonomy 17.92 (exact 17.91666666666666666666666667).
**Conclusion:** Shenzhen is a provisional 2C candidate under SCD v0.2.3, supported by independently operating driverless mobility and the municipal self-healing electricity grid. Both domains show material in-city scale; water sensing and digital government add independent connected and Type 2 capabilities. No proof of three Type 3/D goal-directed service domains exists.

## Source audit
Discovered: 12; examined: 12; accepted: 12; rejected: 0; inaccessible: 0; distinct publisher families: 8. These are this research edition's unique original URLs, not historical Sites search totals.

## Eight-domain decision table
| Bucket | Type winning technology | T,c,g | Type pts | Autonomy winning technology | A,c,g | Autonomy pts |
|---|---|---|---:|---|---|---:|
| MOB | TECH-001 | 2,.65,1 | 5.42 | TECH-001 | 2,.65,1 | 5.42 |
| ENG | TECH-002 | 2,1,1 | 8.33 | TECH-002 | 2,1,1 | 8.33 |
| WAT | TECH-004 | 1,.65,1 | 2.71 | TECH-004 | 0,.65,1 | 0.00 |
| BLD | — | —,—,0 | 0.00 | — | —,—,0 | 0.00 |
| SAF | — | —,—,0 | 0.00 | — | —,—,0 | 0.00 |
| ENV | — | —,—,0 | 0.00 | — | —,—,0 | 0.00 |
| GOV | TECH-003 | 2,1,1 | 8.33 | TECH-003 | 1,1,1 | 4.17 |
| HLT | — | —,—,0 | 0.00 | — | —,—,0 | 0.00 |

**Sum, without per-domain rounding:** 24.79 Type, 17.92 Autonomy. The scoring formula is `100 × Σ_d 0.125 × max_i(g_i c_i T_i/3)` and its analogous A formula.
**Mandatory gates:** Type≥2 domains at T≥2 = ['MOB', 'ENG', 'GOV']; Autonomy C domains at A≥2 = ['MOB', 'ENG']. Result: Type2 True; autonomy C True. Type3 and D gates not demonstrated.

## Technology inventory, evidence and exclusions
### TECH-001 — Pony.ai and Shenzhen Bus Group multi-district driverless passenger services
- **Bucket:** MOB; **status:** operating production; **T=2, A=2, c=.65, g=1**; scope N.
- Coverage evidence: Pony paid robotaxis serving Nanshan and airport/port corridors plus Qianhai/Luohu bus routes with >40k boardings on selected routes; separate operator fleets.
- Human authority / actuation: L4 autonomous vehicles steer, brake and route with safety protocols; some bus services safety operator/limitations. Automatic passenger vehicle navigation and safety decisions.
- Basis and limitations: N provisional: no published Shenzhen-only fleet denominator, geography polygons or total trips. Do NOT assign nationwide 1,975 vehicles to Shenzhen. Sources: SRC-002;SRC-003;SRC-010;SRC-012.
### TECH-002 — Municipality-serving digital power grid self-healing
- **Bucket:** ENG; **status:** operating production; **T=2, A=2, c=1, g=1**; scope W.
- Coverage evidence: Operator claims citywide self-healing system; 91% fault-affected customers can be automatically restored in one minute; all-voltage grid.
- Human authority / actuation: Automated fault detection, isolation and service restoration; strategy-choice demonstration on one 110kV station. Fault isolation, feeder switching and power restoration.
- Basis and limitations: A2 conditional on evidence that dynamic alternatives are actually selected throughout material grid, not just prescribed automatic transfer rules. Sources: SRC-001;SRC-006;SRC-007.
### TECH-003 — Deep Xiao i AI public government assistant and services
- **Bucket:** GOV; **status:** operating production; **T=2, A=1, c=1, g=1**; scope W.
- Coverage evidence: Across municipal services; more than 4 million intelligent Q&As since launch in Feb 2025; 45 unified services in 2025.
- Human authority / actuation: Citizen-facing AI responds, recommends, guides filings; approvals under policy/human governance. Guided service workflows; not open-ended physical action choice.
- Basis and limitations: Do not treat standard one-click official transaction processing as A2. Sources: SRC-004.
### TECH-004 — Municipal online hydraulic and drainage monitoring
- **Bucket:** WAT; **status:** operating production; **T=1, A=0, c=.65, g=1**; scope N.
- Coverage evidence: Online monitoring across 50 primary and 314 secondary drainage divisions; city rain/hydrological stations used for flood warnings.
- Human authority / actuation: Water planners use online warnings and GIS; advanced predictive models currently zone pilot. Data alerts/work-order actions under human process.
- Basis and limitations: N not all customer reach; digital twin in sample basins excluded from municipal T2. Sources: SRC-005.
### TECH-005 — Water predictive twin and Dasha river Harmony program
- **Bucket:** WAT; **status:** pilot/under construction; **T=2, A=1, c=.35, g=0**; scope D.
- Coverage evidence: Advanced modeling in test basins; Dasha all-basin platform launched 2026 but full feature completion future.
- Human authority / actuation: Simulation and optimization advice, not verified independent pump actuation. 
- Basis and limitations: g=0 for municipal scale and unverified operating decisions. Sources: SRC-005;SRC-011.
### TECH-006 — Single intersection fast-brain dynamic signal control
- **Bucket:** MOB; **status:** operating localized pilot; **T=2, A=2, c=.15, g=0**; scope P.
- Coverage evidence: One confirmed Guangming intersection.
- Human authority / actuation: Real-time junction control but no extensive operating street coverage. Automatic phase timing.
- Basis and limitations: Fails citywide N/W gate even if A2 local. Sources: SRC-008.
### TECH-007 — Longhua Avenue time-of-day AI-assisted signal plan
- **Bucket:** MOB; **status:** operating localized corridor; **T=2, A=1, c=.35, g=0**; scope D.
- Coverage evidence: 40 signal lights along 20.3-km corridor.
- Human authority / actuation: Human-engineered plans 24 periods, feedback monitoring. Predesigned signal schedules.
- Basis and limitations: Fails N/W city service scale gate; fixed plans do not prove A2. Sources: SRC-009.
### TECH-008 — Municipal CIM / digital twin and sensor data platform
- **Bucket:** GOV; **status:** operating production; **T=2, A=1, c=1, g=1**; scope W.
- Coverage evidence: CIM includes 4.09m connected sensing streams, 485 documented twin applications; many may be models/not production.
- Human authority / actuation: Supports risk monitoring, simulation and officials, not evidence of automatic integrated multi-domain actuation. Decision support only; local workflows vary.
- Basis and limitations: Same GOV bucket as Deep Xiao i; do not double-count or auto-promote to Level D. Sources: SRC-004.
### TECH-009 — Microgrid and V2G demonstration stations
- **Bucket:** ENG; **status:** pilot; **T=2, A=1, c=.15, g=0**; scope P.
- Coverage evidence: Lianhuashan / Guangming Hongqiao demo sites.
- Human authority / actuation: Demonstration of bidirectional energy transfers in two stations. Test dispatch/transfer.
- Basis and limitations: g=0: two individual facilities; not city-scale autonomy. Sources: SRC-001.

## Why not a higher classification
Type 3 requires three distinct material T3 service domains and a score≥30; not demonstrated. Level D requires three independent goal-directed adaptive domains and score≥30; not demonstrated.
Level C is gated by two materially operating distinct A2 service domains; Shenzhen conditionally meets it, pending verification of power restoration control logic and mobility scale.

## Sensitivity, limitations, editorial check
If grid automatic recovery uses only preprogrammed relay/transfer rather than demonstrable contextual action choice (A2) OR robotaxi/bus service does not reach material N scale, mandatory Level C two-domain breadth fails; autonomy would be at most B. Current score is a conditional research scenario requiring control-logic evidence, audit logs and independent corroboration.
Confirm with switching logs and independent engineering report that Shenzhen grid self-healing selects among dynamic restoration alternatives, not only fixed feeder-transfer rules; measure Shenzhen-only Pony robotaxi fleet, trips, and actual service-zone coverage.

## Source provenance
- SRC-001 [Citywide digital grid with autonomous restoration](https://www.sz.gov.cn/cn/xxgk/zfxxgj/zwdt/content/mpost_13002230.html) — accepted. Says 91% fault-affected users can receive automated restoration within 1 min; citywide fault localization/isolation/transfer.
- SRC-002 [Commercial driverless robotaxi Shenzhen network, 2026 midyear](https://www.sec.gov/Archives/edgar/data/1969302/000110465926098113/tm2623382d1_ex99-2.htm) — accepted. Issuer testimony to paid driverless Shenzhen operations, zones including Nanshan and airport/port; 1,975 fleet figure is all-city total NOT Shenzhen.
- SRC-003 [Operating Qianhai autonomous routes and Luohu shuttle](https://en.szbus.com.cn/news_details/74.html) — accepted. Four Qianhai routes >40k riders and Luohu B888, distinct multi-district operating footprints.
- SRC-004 [2025 top ten digital government smart city developments](https://www.sz.gov.cn/szzsj/gkmlpt/content/12/12608/post_12608373.html) — accepted. Deep Xiao i more than 4m answers; integrated services; 4.09m sensing streams citywide (do not count as 4m independent physical sensors).
- SRC-005 [Digital transformation of municipal water, June 2026 interview](https://swj.sz.gov.cn/gkmlpt/content/12/12870/post_12870647.html) — accepted. Online drainage monitoring 50+314 zones, 284 hydraulic rainfall, 320 meteorological stations, 719 hydrological stations; advanced predictive models only some pilot areas.
- SRC-006 [2025 power-grid AI monitoring](https://www.sz.gov.cn/cn/xxgk/zfxxgj/zwdt/content/post_12000023.html) — accepted. AI equipment anomaly advice with operators checking and acting; not autonomous operations.
- SRC-007 [110 kV self-healing strategy selection field exercise](https://www.cpem.org.cn/list2/105812.html) — accepted. Automatic fault identification, dynamically generated service-recovery strategy and switching in 17 seconds; station-specific corroboration, does NOT alone prove citywide A2.
- SRC-008 [Dual-brain adaptive signal intersection](https://www.sz.gov.cn/cn/xxgk/zfxxgj/bmdt/content/post_12727521.html) — accepted. Dynamic closed-loop traffic timing at ONE junction, excluded from municipal N/W gate.
- SRC-009 [Longhua avenue hourly smart-signal project](https://www.sz.gov.cn/cn/xxgk/zfxxgj/bmdt/content/post_12670975.html) — accepted. 40 signalized intersections along 20.3 km road; area-limited, staff-in-loop plans not citywide Level C.
- SRC-010 [Qianhai commercial/autonomous bus operation](https://qh.sz.gov.cn/sygnan/qhzx/dtzx/content/post_12202672.html) — accepted. Operator/local government description of L4 service and AI road/vehicle/cloud orchestration.
- SRC-011 [2026 open-harmony Dasha river intelligent waters project](https://www.szns.gov.cn/english/news/latestnews/content/post_12702849.html) — accepted. Launch/under development; cannot infer fully operating autonomous water control.
- SRC-012 [Robotaxi 2026 spring paid-order expansion](https://www.sec.gov/Archives/edgar/data/1969302/000110465926034888/tm269906d1_ex99-2.htm) — accepted. Paid Shenzhen orders in Jan/Feb > all 2025; supports real service not merely testing.

## Scope and import rules
This is a research edition. Match existing canonical SCD city records, retain the source/technology/evidence IDs with edition-specific namespacing, preserve earlier versions, and recalculate deterministic scores before publishing. Do not treat search snippets or inaccessible PDFs as examined source material. Do not treat a score of 0 in an unsupported domain as proof of traditional/manual infrastructure.
21 capability tags describe technologies; they do not alter eight independent 0.125 scoring buckets. No inferred duplicate credit across systems.