Analytics and Continuous Control for  Maximum Results, one Micro Change at a Time

Optimize uses deep reinforcement learning (DRL) and automated control to leave the inefficiencies of human-implemented actions behind without relinquishing human supervision.
Fast ramp-up

Unlike standard-issue energy efficiency platforms that need 6-12 months of historical energy operations data to start generating results, Optimize gets started with a simple upload of the blueprints and electrical layout of the data center.  

After upload, Optimize quickly creates a ‘digital twin’ of a customer’s data center, fully simulating the environment to the customer’s exact specifications. Optimize uses its AI agent and APIs—as well as industry-standard communications protocols such as BACnet—to achieve this fast ramp up by seamlessly connecting to widely adopted IoT platforms, data visualization tools, and HVAC systems.

Accurate previews

Optimize trains its AI agent to navigate the physical facility environment.   It can ingest and analyze hundreds of variables from sensor data and regulate the HVAC system settings using as many setpoints and actuators as are exposed.

Within a week, the customer can see both short- and long-term projections of the efficiency gains and cost savings Optimize would deliver.

Lots of small, fast decisions

Moving beyond simulation, Optimize uses its AI agent, APIs, and connections to IoT sensors to ingest, process, and analyze massive amounts of data from the customer’s data center.  

Every few minutes, Optimize makes a full assessment of the data center’s HVAC system – the components, operational status, and environmental interactions – based on data it collects from IoT sensors. Optimize looks at the system holistically and captures the universe of potential actions that can be taken across all of the system’s setpoints and other controls.

Optimize then rapidly projects how each combination of actions will impact future energy usage.  Then, under the supervision of operators, it automatically initiates actions with the local control systems where the changes are verified and implemented.  

This approach results in the best outcomes -- the lowest possible energy consumption while keeping the system within safety, reliability, and other operational parameters.

Users retain control

Optimize automatically grades potential actions on the likelihood that they’ll produce good results.  Operators can choose to limit actions to predefined ranges. Most decisions in Optimize are made automatically, but operators have options to intervene as needed. Essentially, humans frame the decision process and machines handle the implementation.

One key control feature is double-verification.  With double-verification, Optimize checks recommendations against a list of safety and other operational parameters developed by operators.  Once an action clears the list and is sent to cooling systems in the data center, the cooling systems check it against their own mandatory parameters.  If the action clears both checks, it is implemented instantly with no human intervention required, but always with human supervision and control.

Optimize does all this at a scale and with speeds and complexity levels that far surpass human capabilities, even those of the most seasoned facility management staff. And for these human operators, the ’kill switch’ for Optimize is always well within reach.

Watch the Optimize Demo Here!

Data center - Zone 2
Optimized Data Center Simulation
Replay ►






Control Data
With Optimize Agent decision making

Your next step

Want to see what our products can do in your environment? It’s a no-cost, no-risk way for you to see first-hand the energy efficiency gains our products will create for you. 

We would be happy to discuss further. Just complete this form, and we will be in touch to get it scheduled.