In our last post we discussed the challenges of physical buffer vials, in particular (1) the shelf space they take up that could be used for products, (2) the potential to get in the way as product is access around them, and (3) the potential for the buffering materials to spill and ruin product or product packaging. As a solution, TempAlert has developed Virtual Buffer Vials, our patented approach to overcoming the challenges and limitations of using physical buffers.
Using thermodynamic principles, our Virtual Buffer Vials model how a vial of glycol will respond to the temperature of its environment. Instead of using physical buffer vials, we put the measured air temperature of a refrigerator though one of these models to determine what a physical buffer’s temperature would have been.
These temperatures are used as the basis for alerts, which say when a product is approaching the point at which it will no longer be viable or has passed that point and is no longer safe to sell. The data also creates the same kinds of logs that physical buffers would produce, which are required by some state Boards of Pharmacy.
The two main benefits of using virtual buffers are the lack of any physical buffer and the improvements in both accuracy and flexibility of product temperature modeling which can be produced. We have shown that when we simulate a specific type of physical buffer, our models generate temperature estimates that are roughly twice as accurate (i.e. the difference between the modeled and the actual temperature reduces by a factor of two) as taking one CDC-approved physical buffer (e.g. glycol) and modeling it with another CDC-approved physical buffer (e.g. sand). Note: This is taken from unpublished, internal research which is available upon request.
In addition to the benefits of removing the need for a physical buffer, a single air sensor can feed multiple product models, each representing a different product inside the same refrigerated environment. In some cases, two products might be the same solution packaged differently. For example, vaccines come in multi-dose vials or single injection syringes; the volumes are significantly different. In the food industry, think of the different sizes of milk cartons. The larger containers would, in effect, produce more buffering, meaning they would change temperature slower than the smaller containers. Our virtual buffers allow separate models to be developed for different products and run in parallel using the same input air temperatures but producing different modeled product temperatures. In the instance of a power failure where the temperature in the refrigerator environment exceeds the desired range, modeling the two products separately has the potential to recognize that while one exceeded its safety bounds the other did not, allowing more effective risk reduction. It also prevents unnecessary waste if one size is compromised by temperatures out of threshold, while another size was not significantly impacted.
Our suite of virtual product models will steadily increase over time as we work with new customers and expand the range of products we assist them in monitoring for safety. The process of generating a new product model is relatively straightforward and we have demonstrated it for each of the CDC-recommended buffers and for some dairy products.
Transforming remote air temperature readings into alerts about product safety and reports that meet regulatory requirements are just some examples of the value that remote monitoring and IoT can produce.
In future blog posts we will explore other transformations from raw data to actionable and valuable interpretations that further highlight the value that networked sensors can produce.