How Sensors, When Combined With Prescriptive Analytics, Can Assist In Planning For Perishable Goods Supply Chains?
Supply chains of perishable goods are becoming increasingly complex due to competition and broad consumer advocacy about what goes into the production of food (organic vs non organic, use of artificial flavourants and trans fats). Companies are striving to provide high quality products at competitive prices. These companies need to optimise their supply chain to minimise the production cost, delivery cost and the amount of spoiled inventory whilst ensuring various quality standards are met. Risks threatening the supply chain business objectives can be mitigated by collecting data using various sensors at the different stages of the supply chain, analysing the collected data and using this as input into sophisticated analytically based planning in order to make smarter decisions.
Sensors are devices that are used to probe or measure the state of a system. They do so by recording information like voltage, light exposure, temperature, and vibrations among other things. Sensors were traditionally developed for large systems such as airplanes, ships or power stations. However, decreasing sensor prices have made them increasingly ubiquitous. Examples of their use range from autonomous aircraft probing the conditions around them in order to avoid accidents, to an average person using a device to monitor their heart rate while exercising.
The decrease in sensor prices has made it viable to place sensors in stages of production and the supply chain. This leads to a large volume of data being generated that can provide a high definition view of the production and supply chain. In order to be able to reason about how changes in one part of a deeply interconnected production system and supply chain can affect changes in other areas one needs to model the entire system. Modelling the system enables you to plan intelligently based on known and predicted changes in the future state of the operation and importantly to dynamically adapt plans in real-time as changes happen and are measured by sensors.
To illustrate this, let’s consider the example of a company that produces dairy products (e.g. yogurt, fresh cream, butter, etc) and distributes them to retailers. In order for the company to sustainably maximise its profit, it has to take into consideration multiple factors that can affect its operations. Such factors include the demand, manufacturing time, delivery time, spoilage rates under delivery conditions, expiration dates and shelf-life. The company needs to produce sufficient product volumes to cover the demand while at the same time not over-produce as this may lead to losses due to spoiled inventory. Data can be collected in various stages of the operations to optimise processes in real time. For instance, food contamination can be detected at an earlier stage using biosensors to measure the microbial activity. Incorporating this information in the analysis will allow the company to stop the production and limit the damages soon whilst stepping up production at other facilities in order to meet customer demand. Considerable savings can thus be achieved by reducing the amount of raw materials utilised and preventing food and resources wastages.
Delivery time can be predicted by incorporating data generated by telematic/gps devices in the analysis. This allows one to plan effectively given capacity and cost constraints around warehouses, cold storage facilities and delivery vehicles amongst other things. One of the main causes of spoilage associated with perishable goods is temperature variations while on route. The condition of the shipments can be monitored while on route by using sensors to measure temperature variation.
Product flow on the shelfs at the retail store can be estimated by incorporating data collected with barcode scanners. This will make it easy to predict the demand accurately and dynamically as well as allowing one to infer when products are not being effectively restocked on shelves.
By incorporating the various sources of data mentioned above into the analysis, the company will be able to:
- Predict potential problems in advance before they actually happen and plan to accommodate these in the safest and most efficient way possible.
- Handle the impact of unexpected events dynamically in an efficient way.
- Improve the operational planning over time as more data is collected to calibrate the models.
- Use the detailed operational planning for strategic longer-term and capital-allocation decision-making through scenario analysis.
At BusinessOptics, we understand the problems companies face from increasingly complex supply chains. Our visual modelling platform facilitates the modelling of the complexity in manufacturing and supply chains. The flexibility of BusinessOptics allows for customised models to be built rapidly for your business processes. Various parameters such as data from the sensors, business rules and service level agreements can be incorporated in the model to optimise processes and make practical recommendations. These recommendations can easily be applied and integrated into your enterprise systems ensuring maximum automation.
Interested in the BusinessOptics Platform and its associated solutions? Schedule a demo today!