Same routine.
Better timing.

There are 11 named categories and a custom option for other electrical activities. A plan can contain up to 12 tasks. Categories organize your inputs; they do not automatically identify a device or measure its consumption.

You enter operating watts or energy per cycle, duration and quantity. You can also review one of 4 exact appliance references. If the energy is unknown, we ask for it instead of inventing a number.

A timing-only comparison keeps energy, duration and quantity the same. A lower grid CO₂ estimate comes from when electricity is used, rather than a claim that the device uses less electricity.

A valid comparison can show no improvement found. In task planning, you can request a separate 30-minute window extension, then decide whether to apply and review the changed windows. Conflicting or too-short windows have their own errors.

Current carbon guidance uses Great Britain’s national grid signal and Europe/London time. People elsewhere can explore the clearly labeled example, but GB data is not a local recommendation for another country.

No. You choose and operate your equipment. CarbonShift does not connect to a smart meter or start an appliance. Use safe operating times and follow your device’s instructions.

Manual planning works without AI capacity. Optional AI helps interpret your request and explain results, with your review. Ordinary planning needs no wallet; the prototype issuer handles separately consented Sepolia testnet publication.

No. Device energy uses reviewed assumptions and execution is self-reported. Updated grid factors can improve an estimate. A blockchain record can help verify its content and timing; neither establishes measured avoided emissions.