Sharing and Communication around Household Energy Consumption Tawanna Dillahunt Advisor: Jennifer Mankoff HCI Institute Carnegie Mellon University.

Sharing and Communication around Household Energy Consumption

Tawanna DillahuntAdvisor: Jennifer MankoffHCI InstituteCarnegie Mellon University

U.S. households consume over 21.7% of total U.S. energy and generate over 21.1% of total U.S.

carbon emissions [Gardner, et. al, 2008]

Photo Missing

3

Low-Income Households

• 30% of U.S. households make < $30K/year [US Census, 2009]

• Spend greater percentages of income on energy than affluent households [Cooper et al., 1983]

• Median consumption almost as much as affluent households [Shui 2002]

Photo Missing

4

Opportunity

• Low-income individuals are among those more likely to live in rental housing [Belsky and Drew, 2007; McArdle, 2009]

• Renters constitute 30% of U.S. households [Current Housing Reports, 2008]

• Few studies (at the time) targeted low-income households and renters [Chetty, et. al, 2008]

5

Device Limitations

• Will household electricity monitoring devices work within the dynamics of a low-income household?

• What are the dynamics of low-income households in terms of energy consumption?

6

Previous Work

7

Qualitative Studies + Field Deployment

Study 1Energy Use in Low-Income Households [Dillahunt, et. al, Ubicomp 2009]

Study 2Conflicts Between Landlords and Tenants [Dillahunt, et. al, Ubicomp 2010]

Field deployment Sharing and Communication around Household

Energy Consumption

8

• Qualitative study to explore energy consumption in low-income households–Do prior findings generalize to this

community?–Motivations for saving energy?–Existing barriers?

• How can we enhance technology to serve low-income communities?

Study 1

9

Photo-elicitation study

[Clark-IbáÑez, 2004]– Camera– Pen and Notebook

to write about experiences

“Take pictures of objects and/or scenarios that make you think about personal energy use or anything that makes you think

about energy”

Study Design

10

• 26 participants across two locations– 15 NC participants– 11 PA participants

• Diverse payment structures- Pay energy in full- Receive stipend- Pay no energy- Receive allocation

Study Design

11

Findings• Participants received

very little feedback• Saving energy occurred

even if participants did not pay for energy (prior habits)

• Key factors leading to environmental behaviors in low-income households– External barriers– Future generations– Religious beliefs

• Conflict between landlords and tenants around energy consumption

“The faster it [energy meter] spins, the more it costs. The more energy you’re using, the higher your bill is.”

-Angela

Existing Feedback

“I think of, okay, if I keep this

[thermostat] on between 72

and 75, I’m going to have a

low [electricity] bill”

- Erica

Existing Feedback

“This is what they call in our apartments ‘energysavers’. The green light is fine. The red light is what you worry about when that comes on in your

apartment.That means you’re getting alight bill because you are

over. <laughs> And it helps a lot.” - Jacqueline

Existing Feedback

15

Findings• Participants received

very little feedback• Saving energy occurred

even if participants did not pay for energy (prior habits)

• Key factors leading to environmental behaviors in low-income households– External barriers– Future generations– Religious beliefs

• Conflict between landlords and tenants around energy consumption

16

Findings• Participants received

very little feedback• Saving energy occurred

even if participants did not pay for energy (prior habits)

• Key factors leading to environmental behaviors in low-income households– External barriers– Future generations– Religious beliefs

• Conflict exists between landlords and tenants around energy consumption

External Barriers

All photos taken by: Brian

Photos Missing

18

Study 2

• Interviewed landlords to get a balanced perspective

• Story-telling and role play sessions to understand both perspectives

Photos Missing

Sources of Conflict

TENANTS 1PHOTOS +

INTERVIEWS

LANDLORDSINTERVIEWS

TENANTS 2ROLE-PLAYING

Expectations ✔

Money ✔ ✔

Power Imbalance

✔ ✔

Sources of Conflict

TENANTS 1PHOTOS +

INTERVIEWS

LANDLORDSINTERVIEWS

TENANTS 2ROLE-PLAYING

Expectations ✔

Money ✔ ✔

Power Imbalance

✔ ✔

Sources of Conflict

TENANTS 1PHOTOS +

INTERVIEWS

LANDLORDSINTERVIEWS

TENANTS 2ROLE-PLAYING

Expectations ✔

Money ✔ ✔

Power Imbalance

✔ ✔

Sources of Conflict Summary

TENANTS 1PHOTOS +

INTERVIEWS

LANDLORDSINTERVIEWS

TENANTS 2ROLE-PLAYING

Expectations ✔

Money ✔ ✔

Power Imbalance

✔ ✔

Conflict Resolution

TENANTS (1&2)PHOTOS + INTERVIEWS, ROLE-

PLAYING

LANDLORDSINTERVIEWS

Knowledge ✔ ✔

Communication/Negotiation

✔ ✔

Community Action ✔

Conflict Resolution

TENANTS (1&2)PHOTOS + INTERVIEWS, ROLE-

PLAYING

LANDLORDSINTERVIEWS

Knowledge ✔ ✔

Communication/Negotiation

✔ ✔

Community Action ✔

Conflict Resolution

TENANTS (1&2)PHOTOS + INTERVIEWS,

ROLE-PLAYING

LANDLORDS

INTERVIEWS

Knowledge ✔ ✔

Communication/Negotiation

✔ ✔

Community Action

✔

26

Conflict Resolution

STUDIES 1&2TENANTS

STUDY 2LANDLO

RDS

Knowledge ✔ ✔

Communication/Negotiation

✔ ✔

Community Action

✔

27

Solution

Sensing technologies and social computing

can play a role in conflict resolution because of their

abilities to provide new information and better

communication of information

Opportunities

• Sensing technologies produce new information

• Social technologies facilitate sharing• Both technologies influence action

Photos Missing

29

Proposed Work

30

Key Focus Factors

• Sharing led to community action• Better communication helps to

resolve energy-related issues between landlords and tenants

• Negotiation helps to resolve energy-related issues between landlords and tenants

31

Thesis Statement

Eco-visualizations designed to allow individuals tocompare their consumption with others and to actively engage around actions that affect energy consumption will:

• encourage social interaction• raise awareness of energy conservative

behaviors • help residents to negotiate energy use issues

with stakeholders (landlords, housemates, community members)

32

Research Goals

To develop a tool for supporting comparisons and social collaboration

Identify how sharing and collaboration affect energy consumption and communication within communities?

Longitudinal deployment across low-income households of real-time energy monitoring devices

33

Method

34

Speed Dating

• Flesh out usability details

35

Speed Dating Results

• Flesh out usability details

36

Website Design

37

Proposed Study

Longitudinal Deployment

38

Timing

• One year deployment• Checkpoints every 2-4 weeks

39

Thesis Statement

Eco-visualizations designed to allow individuals tocompare their consumption with others and to actively engage around actions that affect energy consumption will:

• encourage social interaction• raise awareness of energy conservative

behaviors • help residents to negotiate energy use issues

with stakeholders (landlords, housemates, community members)

40

Family vs. Community (Option 1)

Website Features Household/Stakeholder Type

Families/Internal(pc/family)

Community/External(kiosk/floor)

Comparison ✓

Social/Discussion ✓

Alerts ✓ ✓

Current Consumption

✓ ✓

Consumption History

✓ ✓

Suggested Actions ✓ ✓

41

Identity vs. Non-Identity (Option 1a)

Website Features Groups

Identity (Group)

Non-Identity (Non-Group)

Comparison ✓ ✓

Social/Discussion ✓ ✓

Alerts ✓ ✓

Current Consumption

✓ ✓

Consumption History

✓ ✓

Suggested Actions ✓ ✓

42

Social vs. Non-Social (Option 2)

Website Features

Targeted Group

Single family household

Single family household

Comparison ✓

Social/Discussion ✓

Alerts ✓ ✓

Current Consumption

✓ ✓

Consumption History

✓ ✓

Suggested Actions

✓ ✓

43

Personal vs. Group Incentive (Option 3)

Website Features Household/Stakeholder Type

Group Individual

Comparison ✓ ✓

Group Incentive ✓

Individual Incentive ✓

Social/Discussion ✓ ✓

Alerts ✓ ✓

. . .

44

Quantitative Measures

• Total energy consumption each month• Number of actions “done” or committed to• Website interaction– How frequently do participants access the

kiosk/pcs?– How long do participants spend interacting with

the kiosk/their pcs?–When do participants access the kiosk/pc?

• Issues reported/issues addressed over time

45

Qualitative Measures

• Pre/Post– Environmental attitudes– Environmental awareness– Attended education event– Did you interact with household

members, neighbors, landlords about the data?

46

Main Contributions

• A tool for supporting comparison and collaboration across households

• Design recommendations for encouraging social engagement around energy consumption across multiple stakeholders

• Demonstrate the usefulness of social computing for ubiquitous computing around energy consumption

• An algorithm for predicting energy consumption based on community or individual conditions ?

47

Schedule

APR FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

- Sample Text

- Sample Text

- Sample Text

On time!

48

Discussion

• Feasibility• What’s more interesting (Option I or

Option II)?• What if no one interacts with the

interventions?• Do I need to add incentives?

49

Thank You

Tawanna [email protected]

Sponsors