E-commerce inventory real-time monitoring system based on Internet of Things

Abstract

The invention relates to the technical field of e-commerce inventory monitoring systems, and discloses an e-commerce inventory real-time monitoring system based on the Internet of Things, which comprises an Internet of Things sensing unit which is deployed in a distributed manner and comprises at least one inventory state detection sensor; the low-power-consumption wide-area Internet of Things gateway is configured to collect and preprocess sensing data; the cloud data processing server is provided with a real-time data analysis module and an inventory prediction model, improves and constructs a three-level early warning system of safe inventory, in-time early warning and abnormal change, combines edge-cloud collaborative analysis (LSTM real-time detection + ARIMA-Prophet root cause analysis) with a 200 + emergency plan library, greatly shortens the stockout and replenishment period, greatly optimizes the early warning accuracy, and improves the early warning efficiency. According to the method and the system, the full-life-cycle data of inventory operation is recorded by using an alliance chain based on Hyperledger Fabric, and multi-party fine-grained sharing (supplier / logistics provider / supervision) and zero-knowledge proof privacy query are supported, so that the dispute handling efficiency is improved, and the supply chain collaboration efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce inventory monitoring systems, and in particular to a real-time e-commerce inventory monitoring system based on the Internet of Things. Background Art

[0002] E-commerce, or e-commerce, generally refers to a new business operation model that enables consumers to shop online, merchants to trade online, and various business activities, trading activities, financial activities, and related integrated service activities in a wide range of commercial and trade activities around the world. It is based on the client / server application mode and allows buyers and sellers to conduct various commercial activities without meeting each other.

[0003] Traditional e-commerce inventory management relies heavily on manual inspections or barcode scanning, which presents problems such as data lag, single dimensions, and lack of environmental perception. Existing IoT monitoring systems generally suffer from technical bottlenecks such as insufficient communication coverage (e.g., a single LPWAN protocol cannot take into account both indoor and outdoor scenarios), low data integration (lack of joint analysis of spatiotemporal dimensions), rigid early warning mechanisms (fixed thresholds cannot adapt to dynamic needs), and equipment battery life limitations (battery power limits deployment density). As e-commerce warehousing develops towards automation and intelligence, there is an urgent need to build a new monitoring system that can achieve real-time perception of multi-source heterogeneous data, intelligent analysis and decision-making, and self-sustaining energy supply. Summary of the Invention

[0004] The purpose of the present invention is to provide an e-commerce inventory real-time monitoring system based on the Internet of Things to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: an e-commerce inventory real-time monitoring system based on the Internet of Things, comprising:

[0006] Distributed IoT sensor units include at least one inventory status detection sensor;

[0007] a low-power wide-area IoT gateway configured to collect and pre-process sensor data;

[0008] Cloud data processing server with real-time data analysis module and inventory forecasting model;

[0009] Multi-terminal visual interactive interface, supporting web and mobile access;

[0010] Intelligent early warning decision module, including multi-level early warning trigger mechanism.

[0011] IoT sensor units are distributedly deployed at various nodes in the warehouse, integrating multiple sensors to realize multi-dimensional data collection (such as commodity existence, weight, spatial location, and environmental parameters). The low-power wide-area gateway serves as a data relay hub, supporting LoRaWAN (long-distance, low-power) and NB-IoT (cellular network) dual-mode communications. Built-in edge computing capabilities enable data preprocessing. The cloud server is the core processing layer, which includes spatiotemporal data fusion, three-dimensional modeling, demand forecasting, and multi-warehouse collaborative optimization modules to achieve global inventory intelligent analysis. The visual interface is a cross-platform interactive terminal (Web / mobile), supporting real-time monitoring and historical data backtracking. The intelligent early warning module triggers early warnings through multi-level thresholds, links to the emergency plan library, and supports strategy self-optimization.

[0012] Preferably, the Internet of Things sensing unit includes:

[0013] RFID reader / writer, used for commodity identification;

[0014] Dynamic weighing sensor, configured in the shelf support structure;

[0015] Three-dimensional spatial sensors using ToF or millimeter-wave radar technology;

[0016] Environmental monitoring sensors include temperature, humidity and light intensity detection components.

[0017] Preferably, the low-power wide-area IoT gateway supports LoRaWAN and NB-IoT dual-mode communication protocols, and has a built-in edge computing module to achieve:

[0018] Data deduplication filtering algorithm based on sliding window mechanism;

[0019] Real-time annotation of abnormal data using the LSTM time series prediction model;

[0020] Adaptive sampling frequency adjustment, dynamic optimization based on inventory change frequency.

[0021] In the dual-mode communication protocol, LoRaWAN is suitable for wide coverage scenarios, and NB-IoT ensures indoor penetration capabilities. The two complement each other to achieve full-scenario coverage. The data deduplication filtering in the built-in edge computing module uses a sliding window mechanism to eliminate duplicate data (such as multiple scans of the same product). The anomaly detection in the edge computing module uses the LSTM model to identify mutation points in time series data (such as a sudden drop in inventory may indicate theft). The adaptive sampling in the edge computing module dynamically adjusts the sensor reporting cycle according to the frequency of inventory changes (such as high-frequency collection during promotional periods). At the same time, it has local cache to ensure continued transmission during network disconnection, device self-diagnosis to monitor sensor status in real time, and AES encryption to ensure data transmission security.

[0022] Preferably, the cloud data processing server includes:

[0023] Spatiotemporal data fusion module, using Kalman filter algorithm;

[0024] Inventory status 3D reconstruction module to generate interactive digital twin models;

[0025] Demand forecasting engine, integrating ARIMA and Prophet hybrid forecasting algorithms;

[0026] Multi-warehouse collaborative optimization module enables dynamic allocation of inventory across warehouses.

[0027] Preferably, the intelligent early warning decision module includes:

[0028] Hierarchical warning trigger conditions, setting safety stock thresholds, expiration warning thresholds, and abnormal change thresholds;

[0029] Multi-channel early warning distribution mechanism, supporting SMS, email, App push and sound and light alarms;

[0030] Emergency response plan library, linking inventory exception types and treatment plans;

[0031] The early warning effect evaluation module uses feedback learning algorithm to optimize the early warning strategy.

[0032] Preferably, the edge computing module is further configured with:

[0033] Local caching mechanism to maintain basic monitoring functions during network interruptions;

[0034] Equipment health self-diagnosis unit, real-time monitoring of sensor working status;

[0035] Secure encryption engine enables end-to-end encrypted data transmission.

[0036] Preferably, the demand forecasting engine further comprises:

[0037] Multi-source data fusion interface to integrate sales data, logistics data and market trend data;

[0038] Real-time incremental training mechanism supports online updates of prediction models;

[0039] Uncertainty quantification module, which outputs the confidence interval of the prediction results.

[0040] Preferably, a blockchain evidence storage module is also included, which is used to:

[0041] Establish an unalterable chain of inventory operation records;

[0042] Realize inventory information sharing among multiple parties;

[0043] Provides an audit traceability interface to support inspection by regulatory authorities.

[0044] Preferably, the IoT sensor unit adopts energy harvesting technology, including:

[0045] Vibration energy harvesters, which use the vibration of storage machinery to generate electricity;

[0046] Light energy collection module with integrated flexible photovoltaic film;

[0047] Adaptive power consumption management system that dynamically adjusts operating mode based on energy reserves.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] First, the present invention builds a three-level early warning system for safety inventory, near-expiry warning, and abnormal changes. It combines edge-cloud collaborative analysis (LSTM real-time detection + ARIMA-Prophet root cause analysis) with a 200+ emergency plan library, greatly shortening the out-of-stock replenishment cycle. At the same time, the early warning accuracy is greatly optimized, and the out-of-stock rate during promotions is reduced.

[0050] Second, the present invention uses a consortium chain based on Hyperledger Fabric to record the entire life cycle data of inventory operations, supports fine-grained sharing among multiple parties (suppliers / logistics providers / regulators) and zero-knowledge proof privacy queries, thereby improving dispute resolution efficiency and enhancing supply chain collaboration efficiency.

[0051] Third, the dual-mode gateway (LoRaWAN / NB-IoT) greatly improves the communication success rate. The spatiotemporal fusion algorithm reduces multi-sensor errors. Combined with vibration / photovoltaic self-powered technology, it extends sensor life, reduces system end-to-end latency, and reduces energy consumption costs per warehouse. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a block diagram of the overall system of the present invention;

[0053] Figure 2 This is a block diagram of the sensor of the present invention;

[0054] Figure 3 This is the low-power wide-area Internet of Things gateway and edge computing and functional block diagram of the present invention;

[0055] Figure 4 This is a block diagram of the cloud data processing server of the present invention. DETAILED DESCRIPTION

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0057] The present invention provides the following technical solutions:

[0058] See also Figure 1 、 Figure 2 、 Figure 3 and Figure 4 , an IoT-based e-commerce inventory real-time monitoring system, including:

[0059] Distributed IoT sensor units include at least one inventory status detection sensor;

[0060] a low-power wide-area IoT gateway configured to collect and pre-process sensor data;

[0061] Cloud data processing server with real-time data analysis module and inventory forecasting model;

[0062] Multi-terminal visual interactive interface, supporting web and mobile access;

[0063] Intelligent early warning decision module, including multi-level early warning trigger mechanism.

[0064] Through the above technical solution, the IoT sensor units are distributedly deployed at each node in the warehouse, integrating multiple sensors to realize multi-dimensional data collection (such as commodity existence, weight, spatial location, and environmental parameters). The low-power wide-area gateway serves as a data relay hub, supporting LoRaWAN (long-distance, low-power) and NB-IoT (cellular network) dual-mode communications, with built-in edge computing capabilities to realize data preprocessing. The cloud server is the core processing layer, which includes spatiotemporal data fusion, three-dimensional modeling, demand forecasting, and multi-warehouse collaborative optimization modules to realize global inventory intelligent analysis. The visual interface is a cross-platform interactive terminal (Web / mobile), supporting real-time monitoring and historical data backtracking. The intelligent early warning module triggers early warnings through multi-level thresholds, links to the emergency plan library, and supports strategy self-optimization.

[0065] IoT sensing units include:

[0066] RFID reader / writer, used for commodity identification;

[0067] Dynamic weighing sensor, configured in the shelf support structure;

[0068] Three-dimensional spatial sensors using ToF or millimeter-wave radar technology;

[0069] Environmental monitoring sensors include temperature, humidity and light intensity detection components.

[0070] Through the above technical solution, the FID reader / writer uses wireless radio frequency identification technology to achieve unique identification of goods and automatic registration of entry and exit. The dynamic weighing sensor is embedded in the shelf support structure, and the inventory level is inferred by weight changes. When the goods are taken away, the weight decreases. The three-dimensional space sensor uses ToF (time of flight) or millimeter wave radar technology to detect the shelf space occupancy rate and the stacking status of goods. The environmental monitoring sensor integrates temperature, humidity and light intensity sensors to ensure that the storage conditions of goods meet the requirements (such as moisture-proofing for food and light-proofing for medicines).

[0071] The low-power wide-area IoT gateway supports LoRaWAN and NB-IoT dual-mode communication protocols, and has a built-in edge computing module to achieve:

[0072] Data deduplication filtering algorithm based on sliding window mechanism;

[0073] Real-time annotation of abnormal data using the LSTM time series prediction model;

[0074] Adaptive sampling frequency adjustment, dynamic optimization based on inventory change frequency

[0075] Through the above technical solutions, LoRaWAN in the dual-mode communication protocol is suitable for wide coverage scenarios, and NB-IoT ensures indoor penetration capabilities. The two complement each other to achieve full-scene coverage, and the data deduplication filtering in the built-in edge computing module uses a sliding window mechanism to eliminate duplicate data (such as multiple scans of the same product). The anomaly detection in the edge computing module uses the LSTM model to identify mutation points in time series data (such as a sudden drop in inventory may indicate theft). The adaptive sampling in the edge computing module dynamically adjusts the sensor reporting cycle according to the frequency of inventory changes (such as high-frequency collection during promotional periods). At the same time, it has local cache to ensure continued transmission during network disconnection, device self-diagnosis to monitor sensor status in real time, and AES encryption to ensure data transmission security.

[0076] The cloud data processing server includes:

[0077] Spatiotemporal data fusion module, using Kalman filter algorithm;

[0078] Inventory status 3D reconstruction module to generate interactive digital twin models;

[0079] Demand forecasting engine, integrating ARIMA and Prophet hybrid forecasting algorithms;

[0080] Multi-warehouse collaborative optimization module enables dynamic allocation of inventory across warehouses.

[0081] Through the above technical solutions, spatiotemporal data fusion integrates multi-sensor data (such as weighing + spatial data) through the Kalman filter algorithm to improve the accuracy of inventory estimation. The digital twin model supports simulating the impact of inventory changes on logistics routes by generating a three-dimensional visualization model of the warehouse. The demand forecasting engine uses the hybrid algorithm ARIMA (time series analysis) to capture periodic patterns and Prophet (Facebook open source model) to handle special events such as holidays. The real-time update incremental training mechanism allows the model to learn the latest sales data online, and outputs the confidence interval of the prediction results through uncertainty quantification (such as "next week's demand: 1000±50 pieces"). Through multi-warehouse collaboration, the prediction results are coordinated and cross-warehouse inventory is dynamically allocated to balance regional supply and demand (such as allocating from warehouse B when warehouse A is out of stock).

[0082] The intelligent early warning decision module includes:

[0083] Hierarchical warning trigger conditions, setting safety stock thresholds, expiration warning thresholds, and abnormal change thresholds;

[0084] Multi-channel early warning distribution mechanism, supporting SMS, email, App push and sound and light alarms;

[0085] Emergency response plan library, linking inventory exception types and treatment plans;

[0086] The early warning effect evaluation module uses feedback learning algorithm to optimize the early warning strategy.

[0087] Through the above technical solution, in the intelligent early warning decision-making module, hierarchical warnings include safety stock thresholds (such as triggering replenishment when inventory is less than 3 days' sales), near-expiry warning thresholds (such as reminding promotions when the product is 30 days away from the shelf life) and abnormal change thresholds (such as initiating a theft and loss investigation when the daily inventory reduction exceeds the threshold). In addition, administrators are notified through SMS / email through multi-channel notifications, and the App pushes them to front-line employees. Sound and light alarms are used for on-site warnings in the warehouse. The emergency plan library can preset processing procedures (such as automatically generating purchase orders when out of stock), associate abnormality types with solutions, and strategy optimization optimizes threshold settings through feedback from the early warning effect evaluation module (such as false alarm rate and response time).

[0088] The edge computing module is also equipped with:

[0089] Local caching mechanism to maintain basic monitoring functions during network interruptions;

[0090] Equipment health self-diagnosis unit, real-time monitoring of sensor working status;

[0091] Secure encryption engine enables end-to-end encrypted data transmission.

[0092] Through the above technical solution, local cache can store data when the network is disconnected and automatically resume transmission after recovery to avoid monitoring interruption. Device self-diagnosis can monitor parameters such as sensor voltage and communication quality, and provide early warning of hardware failures. Secure encryption can ensure that data is encrypted throughout the entire process from collection to transmission to prevent tampering or eavesdropping.

[0093] The demand forecasting engine further includes:

[0094] Multi-source data fusion interface to integrate sales data, logistics data and market trend data;

[0095] Real-time incremental training mechanism supports online updates of prediction models;

[0096] Uncertainty quantification module, which outputs the confidence interval of the prediction results.

[0097] Through the above technical solutions, multi-source data fusion can integrate sales data (historical orders), logistics data (in-transit inventory), market trends (competitive product prices), etc. Real-time training updates the model immediately after new data is generated to avoid prediction lags. Uncertainty output can provide a risk quantification basis for decision-making.

[0098] It also includes a blockchain evidence storage module for:

[0099] Establish an unalterable chain of inventory operation records;

[0100] Realize inventory information sharing among multiple parties;

[0101] Provides an audit traceability interface to support inspection by regulatory authorities.

[0102] Through the above technical solution, this system also includes a blockchain evidence storage module, which is used to establish tamper-proof records. After the establishment, all inventory operations (warehousing, transfer, and outbound) are stored on the chain to form a trusted audit trail. Through multi-party sharing, suppliers, logistics providers, and regulatory authorities can access shared data through permission management, and the regulatory interface provides a standardized API for random inspections by regulatory authorities to support electronic evidence collection.

[0103] IoT sensor units use energy harvesting technology, including:

[0104] Vibration energy harvesters, which use the vibration of storage machinery to generate electricity;

[0105] Light energy collection module with integrated flexible photovoltaic film;

[0106] Adaptive power consumption management system that dynamically adjusts operating mode based on energy reserves.

[0107] Through the above technical solution, in the energy management of the sensor unit, vibration energy collection uses mechanical energy such as forklift operations and shelf vibrations to generate electricity, reducing the frequency of battery replacement. Light energy collection uses flexible photovoltaic films attached to warehouse windows or ceilings to power the sensors. Adaptive power consumption can adjust the working mode according to the remaining power (such as reducing the sampling frequency when the power is low).

[0108] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and alterations may be made to the embodiments without departing from the principles and spirit thereof, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. The real-time monitoring system for e-commerce inventory based on the Internet of Things is characterized by: include: Distributed IoT sensor units include at least one inventory status detection sensor; a low-power wide-area IoT gateway configured to collect and pre-process sensor data; Cloud data processing server with real-time data analysis module and inventory forecasting model; Multi-terminal visual interactive interface, supporting web and mobile access; Intelligent early warning decision module, including multi-level early warning trigger mechanism.

2. The real-time monitoring system for e-commerce inventory based on the Internet of Things according to claim 1 is characterized by: The Internet of Things sensing unit includes: RFID reader / writer, used for commodity identification; Dynamic weighing sensor, configured in the shelf support structure; Three-dimensional spatial sensors using ToF or millimeter-wave radar technology; Environmental monitoring sensors include temperature, humidity and light intensity detection components.

3. The real-time monitoring system for e-commerce inventory based on the Internet of Things according to claim 1 is characterized by: The low-power wide-area IoT gateway supports LoRaWAN and NB-IoT dual-mode communication protocols, and has a built-in edge computing module to achieve: Data deduplication filtering algorithm based on sliding window mechanism; Real-time annotation of abnormal data using the LSTM time series prediction model; Adaptive sampling frequency adjustment, dynamic optimization based on inventory change frequency.

4. The real-time e-commerce inventory monitoring system based on the Internet of Things according to claim 1 is characterized by: The cloud data processing server includes: Spatiotemporal data fusion module, using Kalman filter algorithm; Inventory status 3D reconstruction module to generate interactive digital twin models; Demand forecasting engine, integrating ARIMA and Prophet hybrid forecasting algorithms; Multi-warehouse collaborative optimization module enables dynamic allocation of inventory across warehouses.

5. The real-time e-commerce inventory monitoring system based on the Internet of Things according to claim 1 is characterized by: The intelligent early warning decision module includes: Hierarchical warning trigger conditions, setting safety stock thresholds, expiration warning thresholds, and abnormal change thresholds; Multi-channel early warning distribution mechanism, supporting SMS, email, App push and sound and light alarms; Emergency response plan library, linking inventory exception types and treatment plans; The early warning effect evaluation module uses feedback learning algorithm to optimize the early warning strategy.

6. The real-time e-commerce inventory monitoring system based on the Internet of Things according to claim 3 is characterized by: The edge computing module is also configured with: Local caching mechanism to maintain basic monitoring functions during network interruptions; Equipment health self-diagnosis unit, real-time monitoring of sensor working status; Secure encryption engine enables end-to-end encrypted data transmission.

7. The real-time e-commerce inventory monitoring system based on the Internet of Things according to claim 4 is characterized by: The demand forecasting engine further comprises: Multi-source data fusion interface to integrate sales data, logistics data and market trend data; Real-time incremental training mechanism supports online updates of prediction models; Uncertainty quantification module, which outputs the confidence interval of the prediction results.

8. The real-time monitoring system for e-commerce inventory based on the Internet of Things according to claim 1 is characterized in that: It also includes a blockchain evidence storage module for: Establish an unalterable chain of inventory operation records; Realize inventory information sharing among multiple parties; Provides an audit traceability interface to support inspection by regulatory authorities.

9. The real-time e-commerce inventory monitoring system based on the Internet of Things according to claim 1 is characterized by: The IoT sensor unit uses energy harvesting technology, including: Vibration energy harvesters, which use the vibration of storage machinery to generate electricity; Light energy collection module with integrated flexible photovoltaic film; Adaptive power consumption management system that dynamically adjusts operating mode based on energy reserves.

Images