Refrigeration AI

How it Works

Data Collection

Lizard Wireless Sensors

Our sensors can transmit data for seven years without changing the 2 AA batteries.  They are so reliable that we provide an unlimited warranty while under contract to use our service.

Control System Data

Control System Data is used to help predict refrigeration anomalies and understand why they are happening   If available, we use this information to predict problems and inform users about how to fix them.

Deployment Made Simple

Deploying Lizard sensors has never been easier.  In seconds you can power and place a wireless sensor.

Data Correlation and Threat Analysis

Existing solutions suffer from a variety of problems but one of the most egregious is the problem of 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.

Model pathogen growth

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.

Calculate severity in real-time

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.

Look at the data like a human brain would

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.

White Paper

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.

Designed for Scale

The Lizard platform is feature-rich with robust security, AI enhanced predictions and reporting, search, user tracking, domain specific chat messaging and customer-specific integrations.  Here are some of our most used features.

Reporting & Analytics

Crafted with vital input from customers and designed for scale, ample reporting and research tools allow you to zero in on regions, equipment, and even teams struggling to address systemic issues before they become larger problems.

AI Alert Predictions

Alerts are sent early with AI predictions, in addition defrost cycles, alert pauses, ownership and resolution of each alert are all available on the timeline.


Holistic Alert Settings Across the Enterprise

Tag definitions hold the alert thresholds and communication preferences consistently across all facilities.   Facilities and escalation groups hold the particulars of whom to contact.