Smart Blueprints: Automatically Generated Maps of Homes and the Devices Within Them Jiakang Lu and Kamin Whitehouse Department of Computer Science University.

Smart Blueprints: Automatically Generated Maps of Homes

and the Devices Within Them

Jiakang Lu and Kamin WhitehouseDepartment of Computer Science

University of Virginia

June 20, 2012

“Snap In” Smart Home

2

Motion

Temperature

Light

33

72°F

72°F71°F70°F69°F

“Snap In” Smart Home

4

Motion #1

Temperature #1

Light Switch #1

Motion #2

Sunlight Sensor #1

Light Switch #2

72°F

5

72°F

69°F

6

“Mark and Call” ($2.99 iPhone app store)

[5] Harle, et al

Image from [4] Schindler, et al

Related Work

7

72°F

69°F

Smart BlueprintsInfer floor plan from the app’s sensor data

• No specialized deployment tools or process• Only minor modifications to sensor design

Outline

• Sensor Design• Topological Constraints• Search• Evaluation

8

MotionDoor TemperatureMagnetometer

1111

Outline

• Sensor Design• Topological Constraints• Search• Evaluation

12

1313

14

1515

1616

17

18

EW

19

EW

20

N

S

21

N

S

22

23

N

24

Outline

• Sensor Design• Topological Constraints• Search• Evaluation

25

26

27

28

29

30

31

32

33

Demo

Outline

• Sensor Design• Topological Constraints• Search• Evaluation

34

35

A. B.

C. D.

36

37

38

Downloaded Floor Plans

12.2 average

Limitations

• Failed on 1 out of 4 homes• 1 week to converge• Only evaluated light, motion, mag’r sensors• How many sensors required?• Restricted floor plan encoding

39

43

12

12

3

1 2

43

Conclusions

• Simple sensors are enough to infer layout– Floor plan– Sensor locations

• Requires small modifications to sensor design– Physical pairing– Mounting strategies

• One step towards a “Snap In” smart home40

41

Questions?

(See the demo next!)

42

Backup Slides

43

44

45

Correlation =

𝐴𝐵

46

47

“Snap-In” Smart Home

4848

72°F

72°F

69°F

72°F

70°F

71°F

71°F