Existing solutions suffer from various problems, but one of the most egregious is not modeling pathogen growth and attempting to use rules to create alarms that are then wrong and rarely have valuable actionable information in them. There are several key steps to doing this correctly. In short, model pathogen growth, calculate event severity, understand the root cause, and notify with actionable intelligence.
The growth of pathogens is a multi-phase non-linear function of time and temperature. Unless you take constant automated readings, you don't really know the severity of an incident. There is a direct relationship among temperature, bacterial lag phase and growth rate, in that lag phase decreases and growth rate increases with increasing temperature.
We measure temperature incursions in degree-hours, which is the area under the product temperature curve above (or below) the USDA/EU limits. This can be adjusted to model different products. This has the effect of notifying significant issues early, and minor incursions that do not pose a threat are not notified.
When experts look at data from sensors and control systems they can often instantly tell if an issue requires immediate attention or not, and often what is wrong. This is because they have a vast set of experience to help them. We use a wide variety of data to train AI to recognize complex patterns that simple rules engines will never find. This informs notifications of pending catastrophic failure, and predicts events that require action now providing more reaction time.
A 10-minute read with customer examples and lab results. We start with the advantages of collecting air temperature and computing product temperature in the cloud, all the way to predicting events with AI and adjusting the trade-offs of allowing more reaction time vs. fewer notifications in detail.
Complete food safety visibility across all facilities
Stores can see what’s wrong and take action to save products.
Alerts are sent early with AI predictions. AI knows to prevent false alarms like defrost cycles.
Tag definitions hold the alert thresholds and communication preferences consistently across all facilities. Facilities and escalation groups hold the particulars of whom to contact.
