I. Case Overview

Data source: Dazhong News · Dazhong Daily "Cultural China ｜ 6-Second Face-Swiped Borrow/Return! Jinan Creates Nation's First Unattended Borrowing Closed-Loop Service" (October 19, 2026, reporter Sun Yewen), simultaneously republished by Xinhua News Agency Shandong Channel, Sina Finance, etc. See source list at end of document.

II. Project Background: A Library Receiving 5 Million Annual Visitors, Stuck at the "Finding Books" Step

Jinan City Library receives over 5 million visits annually, peak single-day visits exceeding 30,000, with a collection of approximately 600,000+ volumes. In such a high-traffic public library, one old problem remained unsolved: patrons know the book title but can't find the book.

2.1 Traditional Borrowing Chain: Four Steps, Each Time-Consuming

Before renovation, a complete borrowing required four steps: "search terminal query → note call number → go to shelf area to find → counter or self-checkout scan per book". According to library disclosures, average patron book-finding time was 15 minutes, each borrow/return session took several minutes per person. With tens of thousands of daily visitors, this time was amplified into persistent queuing and crowding.

2.2 Library Management Side: Three Nationwide Common Challenges

1. Low inventory efficiency: High book circulation frequency means book locations change daily; manual inventory is always "yesterday's data".

2. Mis-shelving and dis-shelving: Patrons place books randomly after reading, frequent mis-shelving and dis-shelving mean the next patron goes to the location shown by the system but finds nothing.

3. Labor entrapment: Basic work like inventory, shelf arrangement, and shelf straightening consumes massive labor, trapping staff in low-value repetitive work.

Zhao Xibo, Director of the Network Information Department at Jinan City Library, explicitly stated in the report: "High book circulation frequency, low traditional manual inventory efficiency, and frequent mis-shelving and dis-shelving have become common challenges plaguing libraries nationwide." This sentence points to the project's real value — it's not solving one library's problem, but testing a route that all national peers can take.

2.3 Two Technical Trial Errors: Why the First Two Routes Didn't Work

Notably, Jinan Library didn't start with the right solution. From 2022 to 2024, they tried two popular technical routes, both with unsatisfactory results:

• Inventory Robot: Needed to solve ground navigation, obstacle avoidance, and docking precision issues; actual performance in dense shelf areas did not meet expectations.

• Visual Inventory: Relied on cameras to recognize spine images; recognition rate was unstable when encountering occlusion, lighting changes, dense arrangement, and similar-looking spines.

The "intelligent shelf" solution — embedding antenna arrays in each shelf layer board for real-time on-shelf monitoring — had the best results but was blocked by cost: the high renovation cost made the library hesitate.

So the question was redefined: Is there a device that works well and is also cheap enough? This framing ultimately determined the solution's form.

III. Technical Solution: Unattended Borrowing Closed-Loop Composed of Three Modules

The entire service is interlocked by three modules, solving three things respectively: "countable", "findable", "borrowable fast" — together forming a full closed-loop from inventory to borrow/return.

3.1 Product Layer: Each Book Has a UHF Electronic ID

• Tag Specification: Operating frequency 860–960 MHz, protocol ISO 18000-6C / EPC Gen2, compatible with UHF general reader ecosystem.

• Data Content: Tag contains book ID, call number, etc., mapped to MARC records in the Library Management System (ILS).

• Selection Key Points: Thin tags suitable for inside cover or spine placement; metal shelves and compact shelving environments require attention to antenna detuning, shelf tags can use metal-resistant packaging.

3.2 Collection Layer: RFID Intelligent Cruise Inventory Device (Core Innovation of This Solution)

This is the most noteworthy engineering innovation in the entire project. It didn't follow the traditional approaches of "person carrying device walking" or "each shelf layer installing antennas", but took a third route:

• Track-Based Movement: Tracks laid on top of bookshelves, inventory device automatically slides along tracks daily, no manual pushing needed.

• UHF Antenna Array: Inventory device has built-in UHF RFID antenna array, batch-collects book tags on both sides of shelves while sliding.

• Efficiency Result: 15 minutes to complete data collection and update for 600,000+ volumes, accuracy 98.4%.

• Cost Structure: Overall cost only 3% of intelligent shelf solution — because cost focus shifted from "one antenna per layer" to "one track + one mobile device".

This design approach deserves separate highlighting: intelligent shelves are expensive due to coverage density (need to cover every layer), cruise solutions are inexpensive due to one-time infrastructure (tracks). The larger the collection, the more linear the former's cost rises, the more the latter is amortized — this is why it can achieve a 30x cost difference at the 600,000 volume scale.

3.3 Location Layer: 3D Virtual Library Platform

Inventory data is limited in value if it only says "a book is in the library"; Jinan Library turned it into navigation capability.

• 3D Modeling: 1:1 3D modeling of borrowing area, stereoscopically restored spatial layout, down to each bookshelf and each layer shelf board.

• Precise Location: Patrons click "Virtual Navigation" in WeChat Service Account "Micro Service Hall", enter book title, system returns position and distance precise to "North 1 Area, Shelf 11, Side B, Column 12, Layer 05 on 2nd Floor, 18.83 meters to destination", and plans optimal route.

• Closed-Loop Effect: System doesn't only tell you if a book is in the library — it marks exact location on 3D map, total time from searching to finding approximately 2 minutes.

3.4 Circulation Layer: Unattended Borrow/Return Channel

• Actual Experience: Patron carries 6 books directly through channel, one face-swipe completes borrowing, total 6 seconds.

• Technical Action: During book borrow/return, system instantaneously completes association and disassociation of patron identity and book information, no per-book operation needed.

• Additional Value: Channel simultaneously handles anti-theft function, triggers alarm when un-checked-out books pass through.

3.5 Platform Layer: Integration with Existing Systems

• System Integration: API integration with Library Management System (ILS), completes RFID encoding and MARC record data field mapping.

• Patron Entry: Patron-facing entry is WeChat Service Account, no standalone app installation needed, lowering usage barrier.

• Data Connectivity: Data from five processes — inventory, search, location, navigation, borrow/return — connected on the same platform; this is the meaning of the word "closed-loop".

IV. Route Debate: Why Libraries Started Using UHF Instead of Traditional HF

Libraries are one of the earliest industries worldwide to adopt RFID at scale, but the mainstream has long been HF (13.56 MHz) solutions, typified by ISO 15693 standard book tags. Jinan Library chose UHF (Ultra-High Frequency) as the core for inventory and channel — this is a representative route selection. The real differences between the two frequency bands are as follows:

The tradeoffs in Jinan's solution are clear: place UHF in its two strongest positions — batch inventory and channel gates — while using the 3D virtual library platform to compensate for UHF's inherent shortcoming in single-item positioning precision, and using multi-antenna deployment at channel gates to counter body and occlusion effects. This is not "UHF completely replaces HF", but an engineering judgment of "frequency by scenario". For equipment suppliers, this judgment itself is excellent sales language: don't argue with customers about which frequency band is better — tell them where each frequency band should be used.

V. Four Inventory Technology Route Comparison: Why Jinan Finally Chose the Cruise Solution

Jinan Library actually trialed two routes between 2022–2024; the "measured evaluation" in this comparison table is not extrapolation but conclusions publicly stated by the library.

VI. Patron Perspective: Before vs After Renovation

VII. Implementation Results

7.1 Publicly Measured Data from This Project (Jinan City Library)

All above data comes from public statements by library officials (Network Information Department Director Zhao Xibo and staff Ma Zhengwei) in Dazhong Daily reports.

7.2 Industry Quantitative Reference (Not This Project's Data, For Magnitude Judgment Only)

The following table data comes from other library projects and industry media calibers, has no direct relationship with the Jinan project. When citing, please note respective sources and do not mix with this case:

Note: The "National Library of China" and "National University of Singapore" data groups in the above table are widely relayed in Chinese industry media, but no original source reports were found. When proposing externally, it is recommended to mark as "industry reference" rather than "measured".

VIII. Hardware Configuration List (General Suggestions, Replaceable with Your Company Models)

Jinan Library's official information did not disclose specific equipment models and procurement lists. The following table provides general configuration suggestions for similar UHF smart library projects, can serve as reference framework for quotation and selection:

IX. Pitfall Avoidance Guide: Nine Key Control Points for Library UHF Projects

1. Encoding specification first: Tag encoding standards must be set before starting work. RFID encoding and MARC record field mapping rules must be determined before tagging; otherwise, re-encoding after 600,000 volumes are tagged is an unacceptable cost.

2. Distinguish two types of tags: Thin regular UHF tags are fine for books; shelf tags, because they are pasted on metal layer boards, must use metal-resistant packaging — otherwise detuning results in unreadable tags. This is the only direct metal-resistant tag related part in this project.

3. Unified placement position: Reading consistency differs significantly between inside cover and spine placement; must unify across the entire library and model inventory algorithms around the unified position.

4. Cross-read control: UHF far-field advantage is also a risk — easy to read books from adjacent shelves. Must suppress through reducing transmit power, RSSI threshold filtering, antenna polarization direction design.

5. Precise borrow/return association: Self-checkout counter requires "what's read must be this book on the counter" — UHF far-field is at a disadvantage here. Recommend near-field / short-range antennas with physical shielding on the counter, or keep HF for this step.

6. Body and occlusion: When patrons carry books through the channel, tags are blocked by arms and body. Channel gates must use multi-antenna, multi-angle deployment, and test with real people — don't only test with empty channels.

7. Accuracy expectation management: Inventory accuracy of 98.4% means still approximately 1.6% missed or misread — need to design supplemental inventory process (second verification with handheld devices), don't let staff have unrealistic expectations about the numbers.

8. Privacy compliance: Borrow/return records linked to face information involves personal information protection compliance. Need explicit notice, minimize collection, desensitize borrowing history — public libraries are particularly sensitive to this.

9. Pilot first, then scale: Jinan Library also spent two years of trial and error before finding the solution. New libraries are advised to start with one or two borrowing zones as pilot, run through inventory-location-borrow/return closed-loop before full-library rollout.

X. Replicability and Promotion Prospects

10.1 What Kind of Libraries Does This Solution Suit

• Scale Threshold: The larger the collection, the more obvious the cruise solution's cost advantage over intelligent shelves (cost focus is track infrastructure, amortized by scale).

• Spatial Condition: Bookshelf top has conditions for track laying, or can accept one-time track renovation.

• Pain Point Intensity: Mis-shelving, dis-shelving, hard-to-find books are already concentrated points of patron complaint, sufficient renovation motivation.

• IT Infrastructure: Already has basic ILS system and network environment, capable of API integration.

10.2 Next Steps for Jinan Library

According to library statements, in the future the unattended borrowing technology will be gradually promoted to more public libraries, Quancheng Study Room and other public cultural spaces, building a full-coverage smart reading network. This means this solution has moved from "single library pilot" to "regional replication" phase — for equipment suppliers, this is a market window opening up that can be batch-replicated.

10.3 Three Inspirations for Solution Providers

• Sell experience rather than equipment: What customers are really buying is not readers, but perceptible experience improvements like "patron book-finding time reduced from 15 minutes to 3 minutes".

• Cost narrative first: Jinan Library was deterred by intelligent shelf cost and impressed by the cruise solution's 3% cost — in public cultural sectors, cost structure persuasion often outweighs performance parameters.

• Acknowledge competitor boundaries: Customers may have already trial-errored with robot and visual solutions, coming with failed experience. Proactively clarify each route's boundaries — easier to build trust than merely pushing your own product.

XI. Data Sources and Citation Notes

1. Dazhong News · Dazhong Daily "Cultural China ｜ 6-Second Face-Swiped Borrow/Return! Jinan Creates Nation's First Unattended Borrowing Closed-Loop Service", reporter Sun Yewen, October 19, 2026. Xinhua News Agency Shandong Channel, Sina Finance, Sohu simultaneously republished. (Core data source for this case, including 600,000+ volumes, 15 minutes, 98.4%, 3% cost, 15→3 minute book-finding, 6-second borrow/return, 98% satisfaction, 2022–2024 trial process, etc.)

2. RFID World "Library Management Revolution: How UHF RFID Tags Achieve Second-Level Borrow/Return, Zero-Error Inventory", 2025. (Industry quantitative reference data in Section 7.2 marked as this caliber)

3. Chainway official website case "RFID Library Management in China" (approximately 100,000 volume collection, annual inventory reduced from 3 weeks to 3 hours)

4. GAO RFID, RFIDHY and other supplier public cases (mis-shelving rate, inventory efficiency and other reference data, mostly supplier self-reported caliber, not cited in this document body)

Citation Note: Main case data of this document (Section 7.1) comes from authoritative media direct interviews with the library, directly citable externally; Section 7.2 is industry quantitative reference, has no direct relationship with Jinan project. When proposing externally, must list as separate tables, and note respective sources and calibers to avoid being questioned by professional customers.