Stanford Cleantech Hackathon · Track 1 · Silicon Valley Clean Energy
Be judicious with power.
Judy is a smart plug plus an app. Drag the time into the 4–9 pm peak to watch it work. When something is running that shouldn't be, like a heater in an empty room during peak hours, Judy tells you what it's costing in dollars and CO₂ right now. You tap once to switch it off. After a few taps, Judy learns your pattern and does it for you.
Every number here is computed live from PG&E's 2026 tariff sheets and 12 months of SGIP marginal-emissions data for the PG&E area.
The problem
Time-of-use prices are invisible, and nobody can watch every plug.
The insight
Switching waste off is worth far more than shifting it. Moving a load to off-peak only saves the gap between prices: 3¢/kWh in winter and 12¢ in summer on E-TOU-C. Switching off a device nobody is using saves the whole price, 37–52¢/kWh. Real grid data adds a second reason. In the PG&E area, overnight power is almost as carbon-heavy as the 6 pm peak (about 0.40–0.54 kg CO₂/kWh), so "just run it later" doesn't clean much up. "Turn it off" does.
How it works
See it, price it, nudge, learn, automate.
Try it: every plan, any hour
Pick a device, your rate plan and a time. Judy prices it.
All rate plans
Judy prices every reading on your actual plan.
Exact PG&E total bundled energy rates. All four plans peak from 4 to 9 pm, but the prices around that peak differ a lot, so the same device left on costs different amounts depending on the plan.
Source: PG&E tariff book. E-TOU-C, EV2-A and E-ELEC use Sheet 2, effective June 1, 2026; B-1 is effective March 1, 2026. Not included: the E-TOU-C baseline credit (−$0.0814/kWh on baseline usage), the daily Base Services Charge, and CARE/FERA discounts. These change the bill total, not which hours are expensive. Hourly Flex: Judy accepts any 24-hour price vector, so PG&E's day-ahead prices drop straight in.
Realistic savings
What one household saves in a year.
Solar + battery homes
A 13.5 kWh home battery under California's Net Billing Tariff typically saves $700–1,200/yr and avoids 230–730 kg CO₂/yr by storing midday sun for 4–9 pm. That's from SGIP evaluation data at realistic battery use; Judy doesn't simulate it. Judy adds value on top: it keeps waste from draining the battery during the peak.
Answering SVCE's questions
Every question from the challenge, answered.
Against the judging approach
Across SVCE's territory
What a rollout across all 13 jurisdictions would do.
Schematic of SVCE's service area. Bubble size shows enrolled homes, and color shows summer-peak kW reduced. Click a city.
About 272,000 SVCE accounts, 90% assumed residential, allocated by city population (Census/ACS). Device ownership is assumed: heaters 35%, window AC 20%, smart thermostats 25%, EVs 20%. Capacity is valued at the CPUC 2024 resource-adequacy average of $14.51/kW-month.
Business model
Who actually pays? We tested six answers.
Our pick: a program run with community choice aggregators (CCAs, like SVCE), paid partly on performance
SVCE buys Judy kits for its customers, starting with renters. It pays a small platform fee plus a bonus for every summer-peak kW Judy can verify, and the plugs themselves measure each device. Customers get the kit free and keep 100% of their bill savings.
- SVCE has budget: about $8M/yr for programs.
- SVCE has deadlines: an automated-response option for every customer by July 1, 2027 under the CEC Load Management Standards, plus 25 MW of flexibility by 2030 and double program engagement by 2027.
- Renters are the gap: 61% of Mountain View and 56% of Sunnyvale households rent. They can't install heat pumps or solar, but anyone can plug in a plug.
Unit economics per enrolled home
The performance fee is about 29% of the capacity value Judy creates. On top of its net, SVCE gets an automated-response option for the July 2027 deadline and a direct channel to engage customers. Kit cost assumes 3 metering plugs at about $12 each in volume plus packaging and shipping. Cloud and support cost about $3/home/yr.
The honest caveats
- Heaters and standby power save customers money but add little summer capacity. The capacity value comes from window ACs, thermostats and EV chargers, so SVCE's MW grow with those enrollments.
- Saving energy slightly reduces SVCE's sales. So we pitch on peak capacity, compliance and the mission, not volume.
- Plug hardware is a commodity. Our moat is the pricing engine, the learning and the device-level measurement, with support for any Matter or Kasa plug, plus thermostat and EV APIs.
- Innovation Onramp grants require 2+ years in business, so the first pilot runs with a partner company or as a small staff-approved pilot.
Commercialization pathway
From a hackathon demo to a utility program.
Data and assumptions
Exactly where every number comes from.
Assumptions are marked as such, and the Savings and Scale pages let you change them. These are estimates, not bills.