Open-source technologies to improve supply chain processes for SMEs of rural regions

Open-source technologies with existing systems can be used with existing Enterprise Resource Planning (ERP) systems in companies that are part of the supply chain. With these kinds of technologies, the companies operating in the supply chain can make a step towards digital transformation, allowing them to improve performance. The technologies introduced are MQTT (Message Queuing Telemetry Transport) and Node-Red (Visual programming solution). Integrating MQTT, Node-Red, and existing ERP can create real-time and light architecture for supply chain management.

MQTT is a publish/subscribe, messaging protocol, for minimal devices, unreliable networks and high latency. Its design principles include a lightweight architecture, efficient use of bandwidth, minimal resource requirements, and decoupling of clients through a broker. These design principles help to minimize network bandwidth and device resource requirements. These principles make the protocol suitable for the “Internet of Things” type of connected devices (MQTT.org 2024).

The process starts at the factory where all the production equipment, such as sensors and PLC:s (programmable logic controllers) are. This equipment will send data to a broker, for example, Mosquito or EMQX. MQTT broker is the connection point where the data is sent and fetched. The broker can be hosted on-site or can be outsourced to a third party, depending on the needs, resources, and level of technological knowledge in a specific case. Using the MQTT protocol allows adding new data sources easily with publish/subscribe messaging transport that is used for connecting.

Node-Red is a flow-based low code programming solution for data collection, visualisation, and transformation. Node Red uses visual editor where user can connect nodes and create workflows. These features will make it suitable for integrating different solutions to work and to start prototyping. Node-Red is installed on systems running NodeJS and runs on multiple operating systems, and it is also offered as a docker container. Once Node Red is installed, it runs as a local server-based application which is accessible through a web browser in local network (Node-RED). The server-based solution and browser accessibility makes it also easy to configure and get access over internet, if need be, however this requires some extra measures that are beyond the scope of this article.

ERP is software used to manage processes, for example manufacturing, finance, supply chain etc. This article will not list any specific ERP because several options are available and integrating specific ERP requires specific actions. This article introduces a general principle of how to exploit MQTT, Node-Red and ERP in supply chains.

Production systems and sensors publish data to the MQTT broker for predetermined topics. This is suitable for having structured data transferred with several devices and low bandwidth conditions (Jaikumar, Brindha, Deepalakshmi, Gomathi 2020).

In this example scenario, a garbage truck is disposing the waste to the recycling point. Truck will be weighted at the scale, where the scale publishes to MQTT topic /recycle point/scale/weight, which includes the weight, ID, time stamp, and the type of cargo. The camera systems identify register plates from vehicles and publish them to another topic. This data from the topics must be then combined, validated, and sent to the ERP system for processing.

At this stage, Node-Red is used to set up the MQTT subscribe node to subscribe and listen to the topics sent by the scale and the camera to the MQTT broker. Workflow can be then set up to process the incoming data, for validation purposes this can include, for example, the type of vehicle in relation to the total weight of the vehicle and accurate timestamps. If the equipment publishes data in different standards, unit conversion can be executed at this point, and the data can be enriched with extra metadata related to the equipment, so the ERP system can receive the data in the form of a single transaction. Workflow can be set to check the register plate and the scale information to validate and form a suitable structured data message to the ERPs application programming interface, so that the ERP can interpret this message as a load reception event.

Incoming data can be production events, equipment status events, maintenance information etc. Data transfer from the Node-Red can be through the Representational State Transfer (REST), database writing etc. Several methods are available; this depends on the ERP system that is used. It is important that the structure of MQTT messages and the fields of ERP are defined to be compatible, with Node-Red acting as a transformation layer. In the garbage truck example, ERP records a load receipt event, attaches customer and contract information, and a registration number. Associated device sensors can later supplement this event, when the load is processed further.

The benefits of this architecture come from modularity and real-time data. ERP can receive data without manual input. Processing logic, protocols and ERP Integrations are kept separate, making it easier for changes and reducing dependencies. Adding new sensors or production lines won’t necessarily demand any or significant changes in the ERP systems; only new MQTT topics and Node-Red processing logic are required. Realtime data can be an advantage when maintenance information is processed: if anomalies are detected, alarms can be triggered in the ERP system.

MQTT and Node-Red with ERP system can form an architecture with many customization possibilities, but creating this kind of system requires a fair amount of technical knowledge and an understanding of both OT (operational technology) and IT (information technology). For example, if the devices or sensors do not have built-in capability to send data over the MQTT protocol, a solution must be created that reads the data from the device or sensor and forwards it to the MQTT broker. Integrators must ensure that protocols, data models, and automation logic are aligned so that the system remains effective.

As an example of ongoing applied development in this area, the Interreg Northern Periphery and Arctic (NPA) project Arctic Edge targets improved supply chain operations for SMEs in remote and rural NPA regions by developing and implementing a cloud-based solution built around IoT and MQTT technologies. The project’s deliverables include a flexible digital architecture to support both internal operations and external supply chain interactions and educational materials to accelerate digital adoption. These objectives align directly with the lightweight, real-time MQTT and integration-layer approach.