Aug 25, 2020Australia's Swinburne University of Technology says it has approximately a dozen Internet of Things (IoT)-based pilots under way, each aimed at improving the quality of products companies produce, as well as making supply chains and operations more efficient. The projects include the testing of sensor-based technology and the data it can provide, as well as analytics, alerting and other intelligence extrapolated from that data. The participating Australian companies produce everything from steel to vegemite. This year, the university has partnered with German company Software AG, which provides the software platform and integrates the IoT solutions in order to improve agriculture, manufacturing and other industries.
Among the recent deployments is one carried out by Bega Cheese, a food company that is using the Swinburne and Software AG technology to track the quality of milk produced at 100 dairy farms. The solution, tested at the university's Internet of Things Lab, is now being deployed across these dairy farms throughout the State of Victoria. Bega Cheese is testing the technology to detect the milk's quality, when it was picked up and when it was delivered to the firm's processing site. The cheese company provides such dairy products as cheese, powdered milk and packaged goods, which it exports to 40 countries. It works with dairy suppliers consisting of farmers across Australia, including 100 in Victoria.
The company provides such dairy products as cheese, powdered milk and packaged goods.
Bega Cheese's Adel Salman
Traditionally, the management of such milk quality and shipping data has been performed manually, according to Adel Salman, Bega Cheese's supply chain general manager. The company sought an automated way to ensure that the milk would be picked up from farms and delivered to its own sites for processing in the most efficient manner, he says. In fact, Salman adds, "Full traceability and visibility of the milk, from farm to the manufacturing processing site, enhances product provenance claims."
The Internet of Things Lab's Dimitrio Georgakopoulos
The company also wanted to minimize transportation costs and enhance its supply chain productivity and competitiveness. "The availability, condition and volume of the milk are all key factors that impact the scheduling of milk pickups from farms," Salman explains. "Further, the relative protein and fat content of milk varies, and knowing this ahead of time provides the potential for milk to be allocated in a way that maximizes its value."
Swinburne University and Software AG began working with Bega Cheese in November 2019. The solution they have since developed consists of sensors in milk tanks using a wireless transmission to a cloud-based server. The collected data enables the live monitoring of conditions, says Dimitrio Georgakopoulos, the Internet of Things Lab's director. All sensors for milk monitoring are incorporated in a stainless-steel enclosure that is inserted into each milk tank. Each sensor device captures data such as the level of protein and fat content in the milk, enabling the company to determine product quality and measure the quantity of milk in the tank.
Data is transmitted via narrowband-Internet of Things (NB-IoT), a low-power wide-area network (LPWAN) cellular technology by which sensors transmit information in a daisy chain to a device that then forwards the data via cellular connectivity. Software AG's software platform, known as Cumulocity IoT, receives that information, analyzes the milk's quality and volume, and displays the data for both the producers and Bega Cheese. The quality data can be used for product pricing, while the volume of milk in the tank can automatically prompt pickup scheduling.
The system enables Bega Cheese to monitor the quality of its products.
Proximity, sound and electrical current sensors can identify when a pickup takes place. The sensors dedicated to this process are installed at the loading area and use proximity, electrical current and sound measurements to identify each tanker's arrival and departure. They can also detect when tank-washing activities take place, thereby providing greater visibility into each tank's status, such as when it is cleaned and ready for refilling. The solution employs ambient temperature and humidity sensors to monitor each farm's environmental conditions, Georgakopoulos says, and this data can be compared against the milk's volume and quality in the following days or weeks, to enable better forecasting of production based on weather and climate conditions.
With the collected data, Bega Cheese can view in real time the quantity, temperature and quality of its milk, as well as related forecasts across the entire supply chain—from the farms to the processing site—and it can also monitor pickup events. Milk transport suppliers with access to the software can view the same information as Bega Cheese, receiving pick-plan updates and pickup event alerts. In addition, Georgakopoulos reports, farmers receive alerts related to milk and pickup events.
By providing the software platform for each application, Georgakopoulos says, Software AG enables Swinburne to focus on building sensor-based IoT solutions. Software AG offers IoT solutions and works with researchers and commercial customers alike. "We have a legacy of working very closely with universities," says Tony Drewitt, the company's head of IoT. "Research [often] becomes a reality," he explains, "so we want to support that research—and when they become commercial solutions, we can then support the customers."
Swinburne's research team has been developing IoT solutions since 2010. "In the beginning," Georgakopoulos says, "we were developing our own infrastructures," but the university has more recently been focusing on data analytics, so it turned to Software AG to gain data capture and management capability. Other projects are now under way with manufacturers of composites materials, such as aircraft parts and metal production. These projects focus on supply chain efficiency, plant productivity and product quality.
In food processing, Bega Cheese is testing IoT technology to improve the consistency of its yeast paste, the key ingredient that makes up Vegemite, and to optimize its yield. The solution consists of a variety of sensors to track the product's consistency, as well as machine settings. The IoT solution leverages a refractometer, in addition to data from existing sensors and programmable logic controllers on the machines.
Traditionally, Georgakopoulos says, yeast paste composition was tracked manually. If the proper consistency was not achieved, the product would be reprocessed. With the IoT solution, however, data can be viewed in real time and an automated "recommender system" can adjust equipment settings to ensure the optimal yield is achieved. By reducing the amount of reprocessing, Georgakopoulos says, "We expect to see a 20 percent increase in productivity."
The Cumulocity IoT platform is sensor- and network-agnostic, Drewitt reports, so it can collect data from a variety of sensors via LoRa, LoRaWAN, Wi-Fi or NB-IoT. "That's very good for us," Georgakopoulos says, "because it allows us to focus on inventing things, and not [on] spending a lot of effort in reinventing another IoT platform for sensor data collection." While the software lives in the cloud, Drewitt adds, it could also run on a user's own network. "Connecting the sensors to the platform takes minutes," he says, while Cumulocity is 80 percent ready to go, out-of-the-box; the remaining 20 percent, he notes, consists of customization for a specific deployment.
Software AG's Tony Drewitt
Salman says he expects technology to enhance the company's milk supply chain productivity and competitiveness "with just-in-time milk pickup and processing requirements on a large scale." Such milk quality management could also increase the company's business growth across higher-value dairy products, he adds, noting that IoT technology could increase Bega Cheese's supply chain sustainability by reducing milk wastage and fuel consumption.
The two-and-a-half-year project has been under way for approximately six months so far, Salman reports. "It's still early stages for us," he says, "but we believe the results to date have been promising." From Swinburne's perspective, the university hopes to continue offering IoT solutions through its partnership with Software AG. "The goal is always to make the industry more efficient and more competitive," Georgakopoulos states. "We try to help our partners achieve that goal."
