The Challenge
Helping accelerate how AI supports performance in highly automated warehouses.
The Qubiz team is working closely with NewCold as an enabler of that progress, helping accelerate how AI supports performance in highly automated warehouses. Our work focuses on strengthening the systems behind cooling and decision-making, ensuring operations teams have the clarity, accuracy, and confidence they need to act fast and iterate even faster.
We are proud to contribute our expertise to the projects the NewCold team is leading, and to see technology play its part as another enabler of performance — a partnership built on shared standards, mutual trust, and the belief that progress is always a team effort.
The Strategy
From operational signals to validated, autonomous control.
We strengthen the systems behind cooling and decision-making — so every hour's decision is faster, safer, and easier to trust.
Refining how decisions are made
We refine how setpoints are produced from live operational signals, ensuring every hourly decision tells a complete and reliable story and helping operations move from raw data to validated control actions faster.
Enforcing safe control at scale
We work within a complex ecosystem supporting cooling across multiple warehouses, enforcing hard physical constraints before any setpoint reaches the SCADA layer — reinforcing the foundation operations depend on.
Bridging simulation and live performance
By maintaining and continuously improving the systems behind the Deep Reinforcement Learning model, we help ensure every decision, simulated or live, delivers clarity, precision, and measurable progress.
Connecting operational signals
We enable teams to combine and analyse data across warehouses, energy, and weather inputs in a unified environment — helping operators identify anomalies and understand trade-offs before decisions reach the cooling system.
What We Built
Turning operational data into autonomous decisions
Our focus is on advancing how cooling decisions are generated, validated, and applied. The work centres on refining how setpoints are produced from live operational signals, ensuring every hourly decision tells a complete and reliable story, and helping operations move from raw data to validated control actions faster.
Building the backbone of safe, scalable AI control
Our team works within a complex ecosystem that supports cooling operations across multiple NewCold warehouses. We enforce hard physical constraints before any setpoint reaches the SCADA layer, reinforcing the foundation that operations depend on and giving teams the confidence to act on accurate insights faster.
Bridging simulation and live warehouse performance
We support the operations team in turning complex simulations into meaningful action. By maintaining and continuously improving the systems behind the Deep Reinforcement Learning model, we help ensure every decision, simulated or live, delivers clarity, precision, and measurable progress.
Connecting the dots across operational signals
We enable operations teams to combine and analyse data across warehouses, energy, and weather inputs in a unified environment, helping operators identify anomalies, understand trade-offs, and build confidence in their decisions before they reach the cooling system.
The Results
Progress made possible by a shared standard.
- Hourly decisions. Cooling setpoints generated from live operational signals — each a complete, reliable decision.
- Multi-site. A shared AI control ecosystem supporting cooling operations across NewCold warehouses.
- Safe by design. Hard physical constraints enforced before any setpoint reaches the SCADA layer.
- Unified data. Warehouse, energy, and weather data combined in one environment for faster, more confident decisions.
In Their Words
“Qubiz is a valuable IT partner, helping us solve all of our technical issues and challenges on all aspects: data analytics, integrations, software development.”
Jeroen Klep— Project Manager Data Insights & Integration, NewCold
Frequently asked questions
01What does Qubiz do for NewCold?
Qubiz strengthens the systems behind NewCold's cooling and decision-making — refining how setpoints are produced from live operational signals and turning raw data into validated, autonomous control actions.
02How is safety enforced in an autonomous cooling system?
Hard physical constraints are enforced before any setpoint reaches the SCADA layer, so autonomous decisions always respect the operational limits the warehouses depend on.
03What role does simulation play?
A Deep Reinforcement Learning model is continuously maintained and improved so that every decision — simulated or live — delivers clarity, precision, and measurable progress.
04Which data sources feed the decisions?
Warehouse, energy, and weather inputs are combined in a unified environment, letting operators identify anomalies and weigh trade-offs before decisions reach the cooling system.
