Research
We study how the equipment on customer sites can work with the grid, and publish what we learn.
The problem
Data centers, vehicles and electrified buildings need power faster than the grid can be built out. Part of that capacity already sits on customer sites.
But that equipment speaks dozens of protocols, comes from hundreds of vendors and was never designed to act together.
What we study
- Interoperability: one model for every device, so software reads and controls it the same way, whoever made it.
- Flexibility: how much a site can shift its load, and how fast, without breaking its own limits or its utility’s.
- Homes: how solar, batteries, vehicles and heat pumps behave together, through the Voltra Energy Research Network.
How we work
We test on real equipment and real sites, with national laboratories where a question needs one. Research data is shared only de-identified.
Current work
Data centers can be good
[paper]With Elastic Energy: how flexible data centers can lower electricity costs for existing customers.
Voltra Energy Research Network
[program]How home solar, batteries and EVs behave on the grid, with participants who opt in.
Writing
We rewrote our authorization service in three days
[essay]What it took to replace a core service quickly and safely.
Why we chose Elixir
[essay]Thousands of long-lived connections to unreliable devices, and the runtime built for them.
We are on the cusp of a new era of power: the decentralized grid
[essay]Why coordinating distributed assets is now a communication problem.