Source-linked AI summary
Towards an autonomous industry 4.0 warehouse: A UAV and blockchain-based system for inventory and traceability applications in big data-driven supply chain management
Tiago M. Fernandez-Carames, Oscar Blanco-Novoa, Ivan Froiz-Miguez, Paula Fraga-Lamas
TL;DR
Industrial inventory is often manual and not real-time, motivating an automated, secure, and decentralized approach. The paper designs and evaluates an RFID-equipped UAV system integrated with blockchain, decentralized storage, and smart contracts. In a real warehouse, the system collected inventory data rapidly, estimated item positions from tag signal strength, and was evaluated across blockchain-related scenarios.
Problem
Manual warehouse inventory is costly, time-consuming, error-prone, and usually not performed in real time.
Method
The paper designs and evaluates a UAV-based RFID inventory and traceability architecture using blockchain, decentralized storage, and smart contracts.
Results
The system collected inventory data rapidly in a real warehouse, estimated item positions from tag signal strength, and showed evaluated decentralized-database response distributions across scenarios.
Takeaways & Limitations
The proposed system supports real-time inventory data collection, traceability, decentralized data handling, and automation of selected processes without human intervention.
Takeaways & Limitations
Indoor positioning requires additional consideration of reflections, diffraction, electromagnetic interference, environmental conditions, and reading distance.
Abstract
from arXiv · showhide
In this paper we present the design and evaluation of a UAV-based system aimed at automating inventory tasks and keeping the traceability of industrial items attached to Radio-Frequency IDentification (RFID) tags. To confront current shortcomings, such a system is developed under a versatile, modular and scalable architecture aimed to reinforce cyber security and decentralization while fostering external audits and big data analytics. Therefore, the system uses a blockchain and a distributed ledger to store certain inventory data collected by UAVs, validate them, ensure their trustworthiness and make them available to the interested parties. In order to show the performance of the proposed system, different tests were performed in a real industrial warehouse, concluding that the system is able to obtain the inventory data really fast in comparison to traditional manual tasks, while being also able to estimate the position of the items when hovering over them thanks to their tag's signal strength. In addition, the performance of the proposed blockchain-based architecture was evaluated in different scenarios.
1. Introduction
Industrial inventory is often manual, costly, slow, error-prone, and not real-time, motivating autonomous UAV-based identification and traceability. The paper proposes combining RFID-equipped UAVs with blockchain, decentralized storage, and smart contracts.
- Motivation: UAVs can perform repetitive and dangerous industrial tasks with little human intervention or supervision.The paper frames UAVs as a key technology for smart factories and industrial automation.
- Motivation: Manual inventory is costly, time-consuming, tedious, error-prone, and generally not real-time.The paper identifies automatic, real-time, efficient, flexible, and safe inventory as the ideal target.
- Prior work: Prior UAV inventory systems include barcode scanners, RFID readers, and 3D-LIDAR-based autonomous positioning and mapping.These systems establish the relevant background for UAV-assisted inventory and traceability.
- Contribution: The paper presents the design and testing of an RFID-based drone for inventory tasks as an early contribution in this area.The authors report finding no other literature solution using a similar identification technology.
- Contribution: The proposed system combines UAV-based RFID inventory with blockchain, decentralized storage, and smart contracts.Blockchain stores and validates collected data, while smart contracts can automate business processes without human intervention.
2. Related Work
Related work covers identification technologies, UAV inventory applications, big-data supply-chain analytics, and the system requirements derived from their limitations. The proposed approach is positioned against shortcomings in reading range, line-of-sight, automation, and decentralized data handling.
- System requirements: The solution requires labeling technology, identification technology, a UAV, and supply-management techniques.Together, these elements support warehouse traceability and automated inventory data gathering, processing, and storage.
- Labelling and identification technologies: Barcodes and QR codes require line-of-sight and have limited reading distances, whereas RFID supports reading from several centimeters to several tens of meters.Passive RFID generally does not exceed 20 m, while active RFID can reach 100 m in unobstructed environments.
- Labelling and identification technologies: Smart labels extend RFID capabilities with functions such as event detection, displays, positioning, communications, and environmental sensing.The paper presents smart labels as a more complex stage in label evolution.
- UAV applications: UAV inventory research includes QR-code, UWB, RFID, and UGV-assisted systems, each addressing identification or positioning in distinct ways.A cited QR-code system reports 98.08% precision, while a UWB-based system reports 5 cm indoor location accuracy.
- Big-data supply-chain management: Big-data analytics can support inventory decisions, supply-chain visibility, demand forecasting, and inventory-cost reduction.RFID-enabled shop-floor data have also been processed into logistics KPIs and managerial guidance.
- Analysis of the state of the art: The state-of-the-art comparison identifies shortcomings that motivate the proposed system.Table 2 compares features of existing UAV-based inventory systems with the proposed system.
3. Design of the System
The proposed architecture uses a UAV-mounted reader and onboard computer to collect and process wireless-tag data before transmitting them to a ground station. The ground station can forward inventory information to blockchain and decentralized storage to support security, redundancy, auditing, and automation.
- Communications architecture: A UAV carries a tag reader and single-board computer that collect, process, and wirelessly transmit item data to a ground station.The reader gathers data from tags attached to items, while the onboard computer processes and forwards them.
- Data destinations: The ground station can send collected information to a decentralized remote-storage network or blockchain.This dual-destination design is intended to enhance cybersecurity and support distributed data management.
- Blockchain functions: Blockchain can store collected data or hashes securely and support validation, trustworthiness, third-party access, and decentralized applications.The architecture also supports smart contracts for autonomous process execution.
- Decentralized storage: Decentralized storage provides redundancy by replicating data across multiple nodes and can improve tolerance to attacks on individual nodes.Its security rationale is avoiding reliance on a single storage source.
- Data processing: Both blockchain and decentralized storage manage item identifiers, while an additional software layer processes associations and inventory information.A cyber-physical system may implement this additional processing layer.
4. Implementation
The implementation combines a modular UAV inventory platform with RFID sensing, decentralized storage, and Ethereum-based data insertion. The system links collected inventory records to OrbitDB/IPFS and blockchain hashes for persistence, access, and trust verification.
- 4.1. UAV Implementation: The prototype uses a cost-effective, modular hexacopter designed for adaptable indoor and outdoor inventory applications.Its configuration balances cost, payload capacity, and reliability, with six motors allowing recovery after one motor fails.
- 4.1. UAV Implementation: An onboard RFID reader collects active UHF tag identifiers and sends them through the single-board computer to a ground-station server.The reader was lightened with a foam case, while tag UIDs are first stored locally in JSON and then transmitted over WiFi.
- 4.2. Implemented Architecture: OrbitDB runs over IPFS to persist inventory data in a distributed peer-to-peer database with versioning, content-addressed hashes, and synchronization through IPFS Pubsub.OrbitDB supports log, feed, key-value, document, and counter database types built on an immutable operation-based CRDT.
- 4.3. Inventory Data Insertion and Reading Processes: Ethereum testnets and smart contracts store inventory JSON data and the OrbitDB hash, with Infura, Node.js, and Web3.js connecting application requests to Ethereum.The insertion flow sends inventory data and the OrbitDB hash to Infura, which invokes the smart contract’s setData function and returns a transaction hash.
- 4.2. Implemented Architecture: The implemented architecture gives remote users access to raw OrbitDB and Ethereum data while supporting a backend and frontend for presentation.The implementation simplifies unevaluated cyber-physical-system aspects because the experiments focus on UAV inventory performance.
- 4.3. Inventory Data Insertion and Reading Processes: The architecture supports data trustworthiness through hash-based integrity checks, blockchain traceability, asymmetric cryptography, and protected Ethereum exchanges.Ethereum and OrbitDB also act as timestamping services, while OrbitDB transactions use public/private keys and Infura exchanges use API and secret keys.
5. Experiments
Experiments in a real industrial warehouse assessed UAV inventory speed, RFID signal-strength-based item localization, and decentralized database and blockchain performance. The system collected inventory data quickly, while response latency varied with payload, network placement, and blockchain testnet behavior.
- Experimental setup: 13 RFID tags were deployed in an isolated 50 m ×40 m warehouse area monitored by the drone.The warehouse was approximately 120 m long and 40 m wide; experiments used an isolated subarea for security reasons.
- Inventory time: 78 s versus 116 s: Test 2 collected the inventory faster than Test 1 despite similar departure points and flight paths.The curves were very similar, but Test 2 completed data collection in approximately 78 seconds compared with 116 seconds for Test 1.
- Inventory time: 229 s versus 26 s: Test 3 required the longest inventory collection, whereas Test 4 completed it after the drone departed near the monitored area's center.The results indicate that the drone's departure point was essential for accelerating tag detection.
- Inventory time: Roughly 30% of tags were read within the first 11 s in most tests, 50% of items were inventoried within one minute, and all within four minutes.The early readings occurred as the drone rose from the ground, aided by the reader's range and omni-directional antennas.
- Signal strength monitoring: SSI increased to around −40 dBm near a tag from −50 to −62 dBm farther away, supporting simultaneous inventory and tag-location estimation when drone positioning is available.The paper notes that indoor positioning and environmental factors must be considered for industrial deployment.
- Decentralized database performance: 0.55 s: OrbitDB's worst tested average insertion time was approximately 0.55 seconds, with response delay remaining stable across both networks.The difference between 13 and 10,000 tag IDs was 31.6 ms on the Intranet and 190.5 ms in the remote-cloud case; larger payloads increased latency.
- Blockchain performance: Ropsten blockchain response times ranged from less than 5 s to more than 70 s for almost 100 smart-contract transactions, exceeding OrbitDB delays.The transactions used 9.4 KB payloads, and elapsed time was measured from block inclusion until validation.
6. Conclusions
The paper designed, implemented, and evaluated a UAV- and blockchain-based system for Industry 4.0 inventory and traceability. In a real warehouse, it collected inventory faster than a human operator, supported SSI-based item localization, and evaluated blockchain architecture performance across scenarios.
- Conclusions: The system collected and processed inventory data in real time and sent it to blockchain and decentralized storage for cybersecurity, redundancy, and decentralized applications.Smart contracts also automated certain processes without human intervention.
- Conclusions: The system collected warehouse inventory data remarkably faster than a human operator and could locate items using their RFID tags' SSI.The blockchain-based architecture was additionally evaluated in diverse scenarios.