The Smart Thermostat: Using Occupancy Sensors to Save Energy in Homes Jiakang Lu, Tamim Sookoor, Vijay Srinivasan, Ge Gao, Brian Holben, John Stankovic,

The Smart Thermostat: Using Occupancy Sensors to Save Energy in Homes

Jiakang Lu, Tamim Sookoor, Vijay Srinivasan, Ge Gao, Brian Holben,John Stankovic, Eric Field, Kamin Whitehouse

SenSys 2010Zurich, Switzerland

Motivation

2

43%

State of the Art

3

$5,000 - $25,000Too much cost!

00:00 24:0008:00 18:00

Home

State of the Art

4

55

60

65

70

75

Te

mp

era

ture

(oF

)

Too much hassle! Too much hassle!

Home Home

Setpoint Setpoint

Setback

Home

User discomfort

Energy waste

5

“How much energy can be savedwith occupancy sensors?”

Using Occupancy Sensors

6

00:00 24:0008:00 18:00

55

60

65

70

75

Te

mp

era

ture

(oF

)

HomeHome HomeHome

“Reactive” Thermostat

The Wrong Way

7

00:00 24:0008:00 18:00

55

60

65

70

75

Te

mp

era

ture

(oF

)

Home Home

Shallow Setback

Increase energy usage!

Slow Reaction Inefficient Reaction

00:00 24:0008:00 18:00

Our Approach Smart Thermostat

8

55

60

65

70

75

Te

mp

era

ture

(oF

)

Home Home

Fast reaction Deep setback Preheating

Automatically save energy!

Rest of the talk System Design

Fast Reaction Preheating Deep Setback

Evaluation

9

1. Fast Reaction “Reactive" Thermostat

10

00:00 24:0008:00 18:00

55

60

65

70

75

Te

mp

era

ture

(oF

)

Home Home

Inactivity detector

Active/Inactive

User discomfort

Energy waste

1. Fast Reaction Smart Thermostat

11

Pattern detector

Active/Away/Asleep

00:00 24:0008:00 18:00

55

60

65

70

75

Te

mp

era

ture

(oF

)

Home Home

Detect within minutesWithout increasing false positives

2. Preheating“Why preheat?” Preheat – slow but efficient

Heat pump

React – fast but inefficient Electric coils Gas furnace

12

00:00 24:0008:00 18:00

55

60

65

70

75

Te

mp

era

ture

(oF

)

Home Home

Energy wasteEnergy waste

How to decide when to preheat?

16:00 18:00 20:00

PreheatReact

Arrival Time Distribution

2. Preheating

13

Exp

ecte

d E

nerg

y U

sage

(k

Wh)

3

2

1

016:00 18:00 20:00

Time

Optimal Preheat

Time

16:00 18:00 20:00

3. Deep Setback

14

Arrival Time Distribution

00:00 24:0008:00 18:00

HomeHome55

60

65

70

75

Te

mp

era

ture

(oF

)

Earliest expectedarrival time

Optimal preheat time

Deep setback

Shallow setback

??

Rest of the talk System Design

Fast Reaction Preheating Deep Setback

Evaluation

15

Home #Residents # MotionSensors

#DoorSensors

A 1 7 3

B 1 3 2

C 1 4 1

D 1 4 1

E 2 5 1

F 3 5 2

G 3 4 1

H 2 5 2

EnergyPlus Simulator

Evaluation Occupancy Data

16

Energy Measurements

Energy Savings

17

OptimalReactiveSmart

Smart: 28.8%

Reactive: 6.8%

A B C D E F G H

En

erg

y S

avin

gs

(%)

-10

0

10

20

30

40

50

60

Home Deployments

Optimal: 35.9%

User Comfort

18

ReactiveSmart

Smart: 48 min

Reactive: 60 min

80

A B C D E F G H0

Ave

rag

e D

aily

Mis

s T

ime

(min

)

40

20

60

100

120

Home Deployments

Person Types

Generalization

House Types

19

Zone 1 Minneapolis, MN

Zone 2 Pittsburg, PA

Zone 3 Washington, D.C.

Zone 4 San Francisco, CA

Zone 5 Houston, TX

Climate Zones

Impact Nationwide Savings

save over 100 billion kWh per year prevent 1.12 billion tons of air pollutants

“Bang for the buck” $5 billion for weatherization Our technique is ~$25 in sensors per home

20

Conclusions Three simple techniques, but able to achieve

large savings: 28% on average low cost: $25 in sensors per home low hassle: automatic temperature control

Promising sensing-based solution

21

Q & A

22

Thank you!