Skip to content
CASIENT

Reference installation  ·  anonymised

One house. Fourteen months. Every decision on the record.

A Sydney house in a 2070 postcode, commissioned in stages across a decade and instrumented by us since July 2025. We publish the numbers that flatter us and the ones that do not, because a system asking for this much latitude in someone's home owes them the receipts.

The installation

Group addresses
629
DALI gateways
2
Relay channels
24
Ducted zones
8
Battery
40 kWh
Solar
15 kW
Residents
4
Running since
Jul 2025

What changed

Overrule rate

0.25%

down from 1.9% at month three

Actions the owner reversed within 10 minutes. The only number that measures whether the house was right about what somebody wanted.

Manual recalls

88%

from 41/day to 5/day

Reduction in scene buttons and app interactions per day, measured against the eight weeks before commissioning.

Unattended uptime

99.94%

14 months, no site visits

Appliance availability. Two outages, both power, both recovered without intervention on restoration.

Grid import

31%

summer vs prior summer

Reduction, same tariff and season, driven by deferring loads into the solar peak rather than by using less.

Arbitration, replayed

A captured morning from the reference house, replayed at real cadence. Group addresses are genuine. Every line names the policy that proposed the action and the disposition the arbiter reached.

APPLY / WRITE
A proposal won and reached the bus.
HOLD / DEFER / VETO
A proposal was beaten, delayed, or suppressed by a higher tier.
YIELD
A person disagreed. The house stood down and recorded it.
Decision log  ·  arbitrationReplay

    What is wrong with this evidence

    • This is one house. It is our house, which means the evidence is real but it is not independent, and we will not pretend otherwise until other people’s houses have run long enough to publish.
    • The comparison period for manual recalls is eight weeks, not a year. Seasonal effects are not fully excluded.
    • The grid import figure holds tariff and season constant but cannot control for a household that also became more conscious of its consumption.
    • Nothing here is audited by a third party. When it is, we will say who.

    We would rather publish this list than have you assemble it yourself during a sales conversation. If any of it disqualifies us for your house, that is the correct outcome.

    The next houses will be other people's. That is when this page becomes worth something.

    Become a partner