Especially at this time of crisis when minutes count as the states scramble to obtain critical medical supplies, it’s vitally important to be able to track many thousands of assets, such as masks, gloves, and ventilators, and quickly stage them where they are needed. What’s missing is a flexible, fast, and easy-to-use software system that can be quickly adapted to track these assets in real time and provide immediate answers for logistics managers. An innovative new software approach called “real-time digital twins” running on a cloud-hosted, highly scalable, in-memory computing platform can help address this challenge.
What are real-time digital twins and why are they useful here?
A “real-time digital twin” is a software concept that can track the key parameters for an individual asset, such as a box of masks or a ventilator, and update these parameters in milliseconds as messages flow in from personnel in the field (or directly from smart devices). For example, the parameters for a ventilator could include its identifier, make and model, current location, status (in use, in storage, broken), time in use, technical issues and repairs, and contact information. The real-time digital twin software tracks and updates this information using incoming messages whenever significant events affecting the ventilator occur, such as when it moves from place to place, is put in use, becomes available, encounters a mechanical issue, has an expected repair time, etc. This software can simultaneously perform these actions for hundreds of thousands of ventilators to keep this vital logistical information instantly available for real-time analysis.
With up-to-date information for all ventilators immediately at hand, analysts can ask questions like:
- “Where are all the available ventilators at this moment?”
- “Show me a list of currently available or soon to be available ventilators in my county right now.”
- “What is the average time that ventilators have been in use per patient for each state?
- “Which hospital in a given state has the most unused ventilators?”
- “How many ventilators currently are in repair by make?”
These questions can be answered using the latest data as it streams in from the field. Within seconds, the software performs aggregate analysis of this data for all real-time digital twins. By avoiding the need to create or connect to complex databases and ship data to offline analytics systems, it can provide timely answers quickly and easily.
Besides just tracking assets, real-time digital twins also can track needs. For example, real-time digital twins of hospitals can track quantities of needed supplies in addition to supplies of assets on hand. This allows quick answers to questions such as:
- “Show me the percentage shortfall in ventilators by state.”
- “Which hospitals in a state currently have more than a 25% shortfall (or excess) in ventilators?”
Unlike powerful big data platforms which focus on deep and often lengthy analysis to make future projections, what real-time digital twins offer is timeliness in obtaining quick answers to pressing questions using the most current data. This allows decision makers to maximize their situational awareness in rapidly evolving situations.
Of course, keeping data up to date relies on the ability to send messages to the software hosting real-time digital twins whenever a significant event occurs, such as when a ventilator is taken out of storage or activated for a patient. Field personnel with mobile devices can send these messages over the Internet to the cloud service. It also might be possible for smart devices like ventilators to send their own messages automatically when activated and deactivated or if a problem occurs.
The following diagram illustrates how real-time digital twins running in a cloud service can receive messages from thousands of physical data sources across the country:
How Do Real-Time Digital Twins Work?
What gives real-time digital twins their agility compared to complex, enterprise-based data management systems is their simplicity. A real-time digital twin consists of two components: a software object describing the properties of the physical asset being tracked and a software method (that is, code) which describes how to update these properties when an incoming message arrives. This method also can analyze changes in the properties and send an alert when conditions warrant.
Consider a simple example in which a message arrives signaling that a ventilator has been activated for a patient. The software method receives the message and then records the activation time in a property within the associated object. When the ventilator is deactivated, the method can both record the time and update a running average of usage time for each patient. This allows analysts to ask, for example, “What is the average time in use for all ventilators by state?” which could serve as indication of increased severity of cases in some states.
Because of this extremely simple software formulation, real-time digital twins can be created and enhanced quickly and easily. For example, if analysts observe that several ventilators are being marked as failed, they could add properties to track the type of failure and the average time to repair.
The power of real-time digital twin approach lies in the use of a scalable, in-memory computing system which can host thousands (or even millions) of twins to track very large numbers of assets in real time. The computing system also has the ability to perform aggregate analytics in seconds on the continuously evolving data held in the twins. This enables analysts to obtain immediate results with the very latest information and make decisions within minutes.
In this time of crisis, it’s likely the case that the technology of real-time digital twins has arrived at the right time to help our overtaxed medical professionals.
Note: To the extent possible, ScaleOut Software will make its cloud-based ScaleOut Digital Twin Streaming Service™ available free of charge (except for fees from the cloud provider) for institutions needing help tracking data to assist in the COVID19 crisis.