Executive Summary
China’s urban delivery sector is undergoing a profound transformation. With an estimated fleet of 25 million urban logistics vehicles and a market valued at 1.8–2 trillion yuan, the rapid deployment of autonomous delivery vehicles (ADVs) is reshaping the landscape. By 2025, over 23,000 unmanned delivery vehicles were operating across China’s express delivery networks alone, and the China Post — the country’s largest postal operator — issued a 7,000-unit procurement order, the largest single autonomous vehicle purchase in global history.
The CFLP Smart Logistics Branch report — spanning 55 pages and over 44,000 Chinese characters — provides the most authoritative assessment to date of the commercialization trajectory, business models, deployment data, policy landscape, and competitive dynamics shaping China’s ADV industry. This article distills the report’s key findings into an independent English-language analysis.
1. Market Scale and Growth Trajectory
The Addressable Market
China’s urban distribution logistics (“城配物流”) market is massive and fragmented:
| Metric | Value |
|---|---|
| Total urban logistics market size | 1.8–2 trillion yuan (~$250–280 billion USD) |
| Urban logistics vehicle fleet (end of 2025) | ~25 million units |
| ADV potential market (50% replacement rate) | 1.25 trillion yuan (~$175 billion USD) |
| Express delivery ADV deployment (end of 2025) | >23,000 units |
| Platform-based unmanned transport capacity | >3,000 units |
Vehicle Sales Trends (2021–2025)
The report provides granular vehicle sales data showing the fleet composition:
| Year | Light Trucks (万台) | VAN Series (万台) | Medium Trucks (万台) |
|---|---|---|---|
| 2021 | 211.0 | 45.3 | 8.95 |
| 2022 | 161.8 | 49.8 | 7.32 |
| 2023 | 189.5 | 54.2 | 8.48 |
| 2024 | 190.0 | 58.6 | 8.64 |
| 2025 | 202.4 | 62.0 | 9.17 |
Source: CFLP report, compiled from Ministry of Transport data
The VAN series — the primary target for ADV replacement — has shown consistent growth from 453,000 units in 2021 to 620,000 in 2025, reflecting the expanding last-mile delivery demand. Light trucks remain the dominant category for warehouse-to-store bulk distribution, while VANs serve the small-batch, short-to-medium-distance segment where ADVs are making their deepest inroads.
2. Policy Environment: 300+ Cities with Pilot Programs
One of the report’s most valuable contributions is its comprehensive policy inventory, documenting regulations across all 31 provinces and provincial-level municipalities from 2021 through May 2026. The Appendix lists over 100 policy documents spanning every region of China.
National-Level Framework
Key national-level developments include:
- “人工智能+” Action Plan: Mandates AI-empowered transportation across multiple provinces
- Ministry of Transport coordination: Multi-ministerial joint mechanisms for intelligent connected vehicles
- National standard development: 44 standards issued or in development (see Appendix)
Provincial Policy Density
The policy rollout shows clear geographical clustering:
| Region | Number of Cities/Provinces with Formal Policies | Notable Cities |
|---|---|---|
| Yangtze River Delta | 15+ | Shanghai, Suzhou, Hangzhou, Nanjing, Wuxi, Hefei |
| Pearl River Delta | 10+ | Shenzhen, Guangzhou, Foshan, Zhuhai, Zhongshan |
| Beijing-Tianjin-Hebei | 8+ | Beijing, Langfang, Baoding, Xiong’an, Zhangjiakou |
| Central China | 12+ | Wuhan, Changsha, Zhengzhou, Luoyang |
| Western China | 12+ | Chengdu, Chongqing, Xi’an, Kunming |
| Northeast China | 8+ | Changchun, Shenyang, Dalian, Harbin |
Key Policy Milestones
- Beijing (Feb 2021): First city to issue formal ADV management rules (“无人配送车管理实施细则”)
- Shenzhen (Feb 2025): Issued guidance to expand pilot zones and accelerate functional unmanned vehicle deployment
- Shandong (Sep 2025): Linyi released a comprehensive city-wide unmanned last-mile delivery pilot plan
- Jiangsu (Aug 2025): Provincial-level guidance for unmanned equipment commercial demonstration
- Guangdong (Jan 2026): Provincial AI-empowered transportation development policy
- Inner Mongolia: Ordos became first city to enact a formal regulation (条例) for intelligent connected vehicles and functional unmanned vehicles
Policy Gaps and Challenges
The report identifies a critical bottleneck: the existing Road Traffic Safety Law is built around human drivers and cannot legally define ADV identity. This creates three interlocking problems:
- Licensing: No standardized plate-issuance framework
- Liability: Accident responsibility attribution remains unclear
- Insurance: No dedicated insurance product framework
The report recommends classifying low-speed ADVs as non-motor vehicles based on speed, weight, and function, with corresponding product access standards and ongoing supervision mechanisms.
3. Technology and Investment Landscape
Vehicle Economics: The Cost Advantage
The report provides detailed cost comparison data across multiple scenarios:
Monthly Operating Cost Comparison (per vehicle)
| Scenario | Traditional Manual | Autonomous Vehicle | Savings |
|---|---|---|---|
| Express Last-Mile Relay | ¥6,800/month (1 courier + electric tricycle) | ¥4,200/month (6.5m³ ADV, multi-trip) | 38% |
| Fresh Food Nighttime Delivery | ¥7,500/month (1 driver + mid-VAN) | ¥5,000/month (6.5m³ ADV) | 33% |
| Cold Chain Store Replenishment | ¥8,600/month (1 driver + 4.2m refrigerated truck) | ¥5,000/month (6.5m³ ADV) | 42% |
Unit Cost Per Parcel
| Metric | Manual (Electric Tricycle) | Autonomous (6.5m³ ADV) |
|---|---|---|
| Daily capacity | 2,200 parcels | 2,500 parcels |
| Cost per parcel | ¥0.31 | ¥0.16 |
| Reduction | — | 48% |
Vehicle Cost Breakdown
For the mainstream 5-7m³ ADV class (the most deployed category at 64.6% of the express delivery fleet):
| Component | Cost |
|---|---|
| Bare vehicle (5-year lifespan) | ~¥50,000 |
| Autonomous driving software (annual) | ¥12,000–15,000 |
| Annual maintenance (20-30% of vehicle+software) | ¥12,400–19,500 |
| Monthly total cost of ownership | ¥4,000–5,500 |
This compares to ¥7,000–9,000/month for traditional manned delivery, representing an overall ~40% cost reduction.
China Post’s Mega-Order: Vehicle Mix Analysis
The China Post 7,000-unit procurement — the largest in global history — reveals the vehicle size distribution:
| Vehicle Category | Percentage | Volume Equivalent |
|---|---|---|
| Small (<5m³) | 10.26% | ~718 units |
| Medium (5–7m³) | 64.63% | ~4,524 units |
| Large (8–12m³) | 25.11% | ~1,758 units |
This distribution confirms that the 5–7m³ medium class is the industry’s sweet spot, directly replacing electric tricycles and micro/mid-VANs for last-mile relay transport.
Financing and Investment
The report documents significant capital flows into the sector:
- ZK (九识智能): Won the largest share of the China Post mega-contract as primary supplier for four core packages and first backup supplier for four additional packages. Signed cooperation with the State Post Bureau Development Research Center to establish a “National Post Bureau Unmanned Vehicle Technology R&D Center”
- Neolix (新石器): Cumulative deliveries exceeding 1.5 million, total mileage surpassing 40 million km, daily peak orders of 6,500+. Launched joint “Lightning Island” (来电岛) charging/depot network with TELD (特来电)
- White Rhino (白犀牛): Active in multiple platform partnerships
- Strategic partnerships: Didi Freight + Neolix (Qingdao, 1,200 units); Kuaigou Dache + ZK (Xuzhou); Huolala + White Rhino
Infrastructure Investment: The “Lightning Island” Network
Neolix and TELD jointly announced the world’s first unmanned vehicle automated charging and operations center — “Lightning Island” (来电岛):
- 100 parking spaces with automated charging terminals
- 90 vehicles charged per hour
- High-density vertical parking for urban center deployment
- Integrated automated washing, self-inspection, and maintenance
- 3-year plan: 100 cities, 300 Lightning Islands, 3,000 relay islands, 50,000 charging robots serving 300,000 vehicles
- Global expansion: 10 overseas cities, 30 Lightning Islands
4. Business Model Evolution
The report identifies four distinct business models, tracking a clear migration from asset sales to capacity services:
Model 1: Vehicle Leasing (整车租赁)
Traditional model where operators lease ADVs from manufacturers. Early-stage dominant model, now giving way to more service-oriented approaches.
Model 2: Vehicle Sales (整车销售)
Direct purchase by logistics operators, who own and operate the fleet. Common among large courier companies (SF Express, ZTO, YTO).
Model 3: Transport Capacity Services — RaaS (Robot-as-a-Service)
The fastest-growing model. Platforms integrate ADVs into their order dispatch systems:
- Didi Freight + Neolix (Qingdao, June 2025): Users order ADVs through the Didi Freight app, with on-demand pickup and delivery. Cumulative deliveries >1.5 million, peak daily orders 6,500+
- Kuaigou Dache + ZK + Xuzhou Public Transport Group (November 2025): Joint unmanned freight model using public transport depots as operational bases
- Huolala + White Rhino: Similar platform integration in Wuxi and Suzhou
Current pricing: Services are launching at aggressive ¥9.9 introductory pricing. At full utilization, per-order fulfillment cost can reach 50-60% of traditional transport. However, the report cautions that long-term sustainability of this pricing has yet to be validated at scale.
Model 4: Self-Operated Fleets (运力自营)
Platform-native operators building their own ADV fleets:
- Meituan (美团): Full Shenzhen citywide coverage across 100+ communities, commercial districts, office parks, and campuses. Instant retail delivery (groceries, flash purchases, food delivery) within 30 minutes to 2 hours. Started Saudi Arabia testing (Riyadh, August 2025), with UAE market entry planned
- JD.com (京东): Self-developed vehicles integrated with their logistics network
Competitive Focus Shift
The report identifies a critical strategic pivot: competition is shifting from “whose technology is more stable, whose per-vehicle cost is lower” to “whose capacity network is more complete, whose scaled operations capability is stronger.”
The next phase of competitive differentiation will center on:
- Parking and charging infrastructure integration
- Emergency response and breakdown recovery capabilities
- Maintenance and operational support ecosystems
5. Application Scenarios: Deep Dive
5.1 Express Delivery: Deepest Penetration
Express delivery is the most penetrated ADV application, driven by extreme cost optimization pressure and persistent labor shortages (unstable courier workforce, recruitment difficulties).
Deployment by Major Express Companies (end of 2025)
| Company | Estimated ADV Deployment |
|---|---|
| China Post (中国邮政) | 3,000+ |
| SF Express (顺丰) | 3,000+ |
| ZTO (中通) | 3,000+ |
| STO (申通) | Significant |
| YTO (圆通) | Significant |
| Yunda (韵达) | Significant |
| J&T Express (极兔) | Significant |
| Cainiao (菜鸟) | Significant |
| JD.com (京东) | Significant |
Human-Robot Collaboration Model
In relay scenarios, the emerging model is human-robot collaboration: ADVs transport parcels along fixed routes to designated relay points, where human couriers collect and complete doorstep delivery. Operational data from SF Express, ZTO, and J&T网点 shows:
- Time savings: 1.5+ hours saved per courier per day
- Efficiency improvement: 20-30%
- Per-parcel transport cost reduction: ~40%
Rural Last-Mile Case Study: Yongjia County, Zhejiang
- Challenge: 80%+ mountainous terrain, roads only 3-4 meters wide, tunnels required
- Deployment: 5 ADVs covering 3 towns and 16 administrative villages since April 2025
- Results: “County-to-village same-day delivery” achieved; daily deliveries >3,000 items; complaint rate dropped >50% year-on-year
- Productivity: ADV took over 60% of remote village deliveries; courier monthly collection volume increased 20%
5.2 On-Demand Freight Platform Model
The integration of ADVs with digital freight platforms represents a new mode of innovation. The core value proposition goes beyond cost reduction — it activates incremental market demand by filling supply gaps for fragmented short-haul transport that was previously economically unviable.
Case Study: Yiwu Trading Hub Peak Season
- Context: Yiwu, the world’s largest small commodities trading hub, processed 4.02 billion express parcels in Jan-Apr 2026 alone. During peak season (May-June), daily shipment volumes double
- Pain point: Merchants faced structural supply-demand mismatch — peak season vehicle shortages, off-season idle capacity
- Solution: Neolix RaaS deployed April 2026 — merchants order via mobile, ADVs arrive on-demand, per-order payment
- Result: “Elastic capacity” model enables flexible intra-city shuttling, transforming previously uneconomical short-haul trips into regularized business
Platform Deployment Scale
| Platform | Technology Partner | Launch Cities | Estimated Fleet |
|---|---|---|---|
| Didi Freight (滴滴送货) | Neolix, Cainiao, ZK | Qingdao, Jining, Weifang, Huai’an | ~2,000 |
| Kuaigou Dache (快狗打车) | ZK | Xuzhou | Pilot scale |
| Yunmanman (运满满) | Neolix | Wuxi, Suzhou | Pilot scale |
| Huolala (货拉拉) | White Rhino | Various | Pilot scale |
Total platform-based unmanned capacity: >3,000 vehicles
5.3 Planned Logistics: The Breakthrough Frontier
Planned logistics accounts for >80% of urban distribution volume (including small-B customers) — characterized by scheduled, standardized, multi-point distribution with value-added services. This is the largest addressable segment and the next frontier for ADV penetration.
Substitution Readiness Assessment
| Scenario | Batch Size | ADV Readiness | Key Barriers |
|---|---|---|---|
| Express last-mile relay | Small, high-frequency | ✅ Active deployment | — |
| Convenience store replenishment | Small-medium, multi-point | ✅ Active deployment | Unloading, unmanned handover |
| Fresh food nighttime delivery | Medium, fixed schedule | ✅ Growing | Temperature control, timing |
| Industrial short-haul transport | Medium-heavy, fixed route | ✅ Growing | Load capacity, stability |
| Supermarket bulk delivery | Large (0.5-3 tons/store) | ⚠️ Challenging | Unloading, shelf-stacking, container returns |
| Hospital/medical delivery | Small, high-value | ⚠️ Challenging | GSP compliance, chain of custody, specialized services |
| Production material JIT | 3-10 tons, strict timing | ❌ Not yet viable | Zero-tolerance for delays, production-line integration |
Cainiao + Laiyoupin Retail Replenishment Case
- Cainiao deployed 10+ ADVs in Hefei for warehouse-to-store delivery to Laiyoupin snack chain stores
- Coverage: stores within 50km radius of warehouse
- Average delivery time: 1.5 hours (same-day warehouse-to-store)
- Impact: replenishment frequency increased from once every 2 days to 1-2 times daily
- Store inventory cost reduction: ~30%
- Expansion: rolled out to 30+ cities across Chengdu, Chongqing, Wuhan, Ganzhou, Xiangtan, Zhuzhou
5.4 Nighttime Delivery
ADVs are uniquely suited for nighttime delivery — a scenario that is policy-friendly (off-peak traffic, emissions reduction) but faces severe human labor challenges (night shift recruitment difficulties, driver health impacts).
Case Study: Braised Food Chain Nighttime Delivery
- Challenge: Cross-city early-morning ingredient transport with stringent freshness and punctuality requirements; nighttime driver recruitment near-impossible
- Solution: ZK L4 autonomous vehicle deployed for fully unmanned nighttime transport — vehicle departs autonomously in the evening for the slaughterhouse, loads automatically at the processing plant, and returns via autonomous navigation
- Results: Transport cost reduced 8%, fuel cost reduced 10%, overall monthly operating expenses down 5-10%
5.5 Industrial Short-Haul Transport
Short-haul industrial transport — moving materials between warehouses, production lines, and staging areas — has emerged as a strong ADV fit due to its fixed routes, high frequency, and standardized requirements.
Case Study: Nantong Xingchen Electronics Factory
- Requirement: Heavy-load, high-frequency short-haul between two factory sites (16km route)
- Cargo: Aluminum, paper, high-density bulk platinum and other raw materials
- Frequency: 4-5 round trips daily
- Solution: ZK L5 RoboVan with L4 autonomous driving + deep forklift integration, 1.8-ton maximum load capacity, compatible with public DC fast-charging infrastructure
6. Standards and Regulatory Framework
The report’s Appendix catalogs 44 standards issued or in development, spanning national, industry, local, and group standards:
| Category | Count | Examples |
|---|---|---|
| Industry Standards (YZ/T) | 2 | Unmanned vehicle mail/parcel delivery service specifications; Technical requirements for delivery unmanned vehicles |
| Local Standards (DB) | 5 | Beijing, Chongqing, Suzhou — vehicle testing methods and terminology |
| Group Standards (T/) | 37+ | Covering safety requirements, operational management, testing procedures, personnel competency, networking, campus delivery, etc. |
The standards ecosystem reflects the rapid maturation of the industry but also the fragmentation challenge: with standards being developed at multiple levels by multiple bodies, the report calls for greater national-level coordination.
7. Challenges and Bottlenecks
The report identifies five interconnected challenge categories:
7.1 Policy and Regulation (致命级 / “Fatal-Level”)
The most critical barrier. The Road Traffic Safety Law’s driver-centric framework cannot accommodate autonomous vehicles. Three gaps need resolution:
- Identity classification: ADV legal status undefined
- Road access rights: No unified licensing framework
- Safety supervision: Fragmented multi-department coordination
Recommendation: Classify low-speed ADVs as non-motor vehicles; establish multi-level monitoring platforms following the “enterprise monitoring → government supervision → networked joint control” model used for road transport vehicles.
7.2 Technology Reliability
While ADVs can handle standard open-road operations, long-tail scenarios remain problematic:
- Erratic e-bike behavior
- Temporary road obstructions (construction materials)
- Unstructured mixed traffic (pedestrians + vehicles)
- Adverse weather conditions
The report calls for dual optimization: rigid rule compliance balanced with flexible scenario adaptation, and operational efficiency balanced with impact on existing traffic systems.
7.3 Safe and Stable Operations
Real-world challenges include:
- Conservative interaction logic causing traffic congestion
- Vulnerability to malicious stops/interference
- Interaction capability with existing traffic participants
7.4 Infrastructure Mismatch
- Charging infrastructure still sparse outside pilot zones
- Depot and parking facilities not designed for autonomous fleets
- Road infrastructure (markings, signage) quality varies significantly
7.5 Public Trust and Social Acceptance
- Public skepticism about “driverless” safety
- Employment displacement concerns
- Data security and privacy risks
8. Commercialization Roadmap: 4-Phase Evolution
The report outlines a clear four-phase trajectory:
| Phase | Timeframe | Characteristics |
|---|---|---|
| Phase 1: Technology Validation | Pre-2024 | R&D and product iteration; closed-park testing |
| Phase 2: Commercial Exploration | 2024-2026 | Vehicle deployment, operational system building, business model experimentation |
| Phase 3: Scaled Replacement | 2026-2030 | Ecosystem operations, scaled profitability, integration with urban logistics systems |
| Phase 4: Near-Full Replacement | 2030+ | Cross-ecosystem operations, seamless trunk/branch/last-mile integration |
The workforce evolution mirrors this timeline:
- 2026: Human-dominant, early human-robot coexistence
- 2030: Full human-robot symbiotic urban logistics
This translates to a workforce transformation from “driver” → “safety monitor” → “fleet supervisor” → “capacity architect,” with new roles emerging: ADV safety officers, maintenance technicians, dispatchers, and remote operations specialists.
9. Competitive Landscape: Key Players
Technology Providers (Vehicle + Autonomous Driving)
| Company | Positioning | Key Metrics |
|---|---|---|
| ZK (九识智能) | Full-stack L4/L5 provider | Won China Post mega-order (primary supplier, 4 packages); State Post Bureau R&D Center partner; L5 RoboVan with 1.8-ton payload |
| Neolix (新石器) | RaaS pioneer | 1.5M+ deliveries, 40M+ km, 6,500 daily peak orders; Lightning Island network with TELD; Didi Freight partnership (1,200 units) |
| White Rhino (白犀牛) | Multi-platform partner | Active with Huolala and multiple freight platforms |
| Meituan (美团) | Self-operated fleet | Full Shenzhen coverage, 100+ communities; Saudi Arabia testing started; instant retail focus |
| JD.com (京东) | Self-operated fleet | Integrated with JD Logistics network |
Platform Enablers
| Platform | Role | Fleet/Partnerships |
|---|---|---|
| Didi Freight | Traffic hub + dispatch + settlement | Neolix, Cainiao, ZK; ~2,000 vehicles |
| Kuaigou Dache | Platform integration | ZK + Xuzhou Public Transport |
| Huolala | Platform integration | White Rhino |
| Yunmanman | Platform integration | Neolix |
10. Future Outlook: Key Projections
Market Size
| Projection | Value |
|---|---|
| Current urban logistics vehicle fleet | 25 million units |
| 50% replacement rate target | 12.5 million ADVs |
| Potential market value at replacement | 1.25 trillion yuan (~$175 billion USD) |
| Phase 3 (2026-2030) | Scaled profitability, ecosystem-level operations |
Penetration Trajectory
The report envisions ADVs becoming the primary transport capacity for urban distribution rather than merely a supplementary force:
- Short-term (2026-2028): Single-scenario, single-link supplementary capacity
- Medium-term (2028-2030): Multi-scenario, multi-link primary capacity
- Long-term (2030+): Full-scenario integration forming a citywide intelligent logistics network
Industry Structure Implications
- From asset ownership to capacity access: Logistics companies shift from owning fleets to accessing unmanned capacity on-demand
- Platform consolidation: Freight platforms become the central nervous system — traffic hub, dispatch brain, settlement engine
- Infrastructure as competitive moat: Charging, parking, and maintenance networks become the new barriers to entry
- Standard export: Chinese ADV standards and operational models are already going global (Neolix’s 10-country Lightning Island plan, Meituan’s Middle East expansion)
Workforce Transformation
The report emphasizes that the transition must be managed:
- Traditional drivers can transition to ADV safety officers, maintenance technicians, and dispatchers
- New roles: Remote operations specialists, fleet coordinators, capacity architects
- Education: Calls for university cross-disciplinary programs (autonomous driving + smart logistics + AI) and vocational training systems
Conclusion
China’s autonomous delivery vehicle industry has decisively moved from laboratory validation to commercial-scale deployment. The CFLP report paints a picture of an industry at an inflection point: over 23,000 vehicles in operation, deployment across 300+ cities, a comprehensive (if fragmented) policy framework, proven 40% cost advantages, and a clear commercialization roadmap extending to 2030 and beyond.
The critical challenges — unified legal identity, long-tail technical reliability, infrastructure buildout, and public acceptance — are significant but actively being addressed. The China Post 7,000-unit order signals that the industry’s largest customers are voting with their procurement budgets.
For global observers, this report offers a unique window into the scale and speed of China’s autonomous logistics transformation — a development that will reshape not just domestic logistics but, as Chinese ADV manufacturers execute their global expansion strategies, the future of urban delivery worldwide.
