Direct-to-consumer deliveries are the way of the future, and becoming even more important due to COVID-19. However, logistics companies are struggling to meet demand amid a barrage of factors, from ever-changing road conditions to managing temperature-sensitive deliveries of medications. A powerful new streaming data software technique, called real-time digital twins, can provide the insights needed to answer critical logistics questions, reduce unknown variables, and improve the efficiency of operations.
With the sheer volume of streaming data coming from fleets, smart warehouses and other telemetry sources, it’s a daunting challenge for logistics managers to extract critical insights and identify potential problems in real time. Traditional streaming analytics approaches, such as open-source Apache Storm and Flink, help managers to extract patterns in aggregated message flows. Still, they can’t put this information in context to assess its significance and take effective, individualized action for each data source.
These solutions were designed to pass incoming telemetry through a software pipeline to extract patterns of interest, gain aggregate insights, and send alerts when specific conditions are met. However, they do not track the dynamic state of the data sources themselves, or make inferences about their behavior.
For example, streaming analytics can detect high oil temperature readings in a fleet of trucks, but can’t explain why these readings are occurring, whether they indicate an impending failure, and what specific actions should be taken. To do this, the streaming analytics platform needs to maintain contextual information about each truck, so that it can provide deeper introspection on incoming telemetry and make more strategic decisions about alerting and intervention. For example, one truck engine might be expected to exhibit higher-than-normal oil temperature because of its age and maintenance history, while another, newer engine might be displaying an unusual problem that needs immediate attention.
To keep complex logistics systems running smoothly, streaming analytics needs to provide this deeper level of introspection, especially when receiving telemetry from large numbers of data sources, such as thousands of trucks in a fleet or pallets in a warehouse. Enhanced analysis of this telemetry that makes use of context about each data source can do a much better job of identifying and predicting inefficiencies, potential problems and key trends. It can help answer questions like:
Is a vehicle stopped because it’s at a rest stop or because it has stalled?
Is an engine parameter on the vehicle abnormal, or is this expected given the vehicle’s known issues and maintenance history?
Will the refrigeration compartment’s current temperature and trend create a problem for the specific medications it currently holds?
Has the driver been on the road for too long with respect to legal requirements?
Does the driver appear to be lost or entering a potentially hazardous area?
The new software technique for streaming analytics called real-time digital twins can help provide the insights needed to answer these questions. It creates a software-based twin of each physical data source being tracked, containing contextual information about the data source (such as the expected parameters and maintenance history for a truck engine). The digital twin hosts application code that analyzes incoming messages from its data source with immediate access to this context, and it continuously updates the context with each incoming message as it “learns” more about the data source’s dynamic condition. Application code in the twin only needs to focus on a single data source instead of managing the flow of all incoming messages, and this enables better feedback for each data source.
A real-time digital twin can run on public clouds, such as Microsoft Azure, which have the scalability to process messages and maintain real-time digital twins for thousands of trucks, smart warehouses or other telemetry sources from across wide geographical areas. The cloud service can also continuously aggregate and visualize key information extracted from all the real-time digital twins to detect emerging issues and boost overall situational awareness for managers. This helps them maintain the big picture and more quickly create strategic responses to major challenges, such as weather delays, highway blockages and power outages.
Here are four examples of situations in which real-time digital twins can help logistics managers dramatically boost the effectiveness and timeliness of their delivery systems:
Changing conditions. Shifting traffic patterns, accidents and even record-breaking hurricanes, storms and fires are causing drivers’ routes to change continuously. With logistics companies tracking thousands of vehicles on the road, real-time digital twins can help analyze how situations like freeway closures due to forest fires will affect each vehicle and alert the drivers to new routes.
Spoiled food and medications. In the COVID-19 era, many high-risk people are seeking alternatives to shopping in crowded stores and now require food and medications to be delivered to their doors. To complete these deliveries, a vehicle may contain hundreds of different temperature-sensitive foods and medications. Real-time digital twins can monitor needed temperatures for each item and alert drivers and logistic managers to issues that threaten the safety of cargo.
Emerging mechanical issues. Maintaining expensive commercial vehicles is complex, involving numerous parameters, such as tire pressure, fluid levels, engine systems and much more. Real-time digital twins can track each vehicle’s mechanical status and alert to needed maintenance, avoiding unexpected delays and costly repairs.
Lost or erratic drivers. Commercial drivers operate under strict rules, with substantial penalties for violations. Real-time digital twins can spot erratic driving behavior or impending infractions, and signal the driver before they occur. They can also detect whether a driver appears to be lost, so that dispatchers can quickly correct the problem and save time and fuel.
The year 2020 has seen a dramatic increase in the use of logistics systems that keep the goods we need flowing to their destinations. Insights gained from real-time digital twins can help logistics companies track the countless components in their networks, keep them operating smoothly, and avoid unnecessary delays. These insights will enable companies to lower their costs and boost on-time performance, so that they can compete effectively in a critical industry on which we all depend.
RSS
