Presentation at SFU Exchange

n May 6th, I presented my research at SFU Exchange, a showcase of graduate student research, ideas, innovations, and entrepreneurial skills. I presented a brief overview of the research problem I am working on, covering some of the psychological research, a summary of available tools, and our information ecosystem approach to designing integrated feedback systems.

My slides and accompanying notes follow:

SLIDE 1: Introduction
My name’s Johnny Rodgers. I’m a Masters of Science student at the School of Interactive Arts & Technology.

I’ve been working with Lyn on these projects for the last 18 months alongside others in our research group. My contributions have been in design, programming, system installation and testing, and research.

I’m going to give a bit more detail on the research surrounding our approach, and talk about a few examples of the tools we’ve built and what we’ve learned in the process.

SLIDE 2: Introduction
My research aims to make the consumption of residential resources visible in order to support informed decision-making.

Problem: We don’t have the information we need in our daily activities to reduce our consumption of energy, water, and other household resources.

Ask someone on the street how many kWh of electricity they used yesterday, and you’ll likely get a blank stare, even though the same person might be able to tell you the mileage they get in their car.

We just don’t have access to this information, except for a hydro bill every couple of months.

Slide 3: Engaging Residents
Research has consistently shown that providing feedback can reduce energy consumption by 5-20%.

However, we can’t just throw information at the problem. As Lyn mentioned, we all already have more information than we know what to do with.

We need to contextualize feedback so it makes sense to people acting in a given situation: in our case, residential activities.

Slides 4-8: Tools
Lots of tools exist to try to address this problem both in the marketplace and research community. They use a multiplicity of approaches:

  • Dashboards & Analytic Interfaces
    - detailed interfaces, task-oriented, high attention
  • Point-of-consumption Feedback
    - at-a-glance, situated
  • Mobile Tools
    - on-the-go, awareness and connection to household
  • Smart Appliances
    - task-specific, opportunities for change, introduce automation into household activities
  • Ambient & Artistic Displays
    - peripheral, subtle, promote curiosity and playfulness

Each helps residents make certain kinds of decisions and answer certain questions.

Slide 9: Motivating Change
The psychological literature helps us understand some aspects of the problem of motivating behaviour change.

Feedback alone is not enough to achieve long-term change. Must be combined with other factors.

Utilities have experimented with various pricing penalty and reward schemes. Perhaps unsurprisingly, penalties are more effective than prizes. However, intrinsic incentives work better than either. That is, if you can achieve buy-in on an individual basis, you don’t need to use carrots or sticks.

Social influences in the form of public commitments and peer champions are particularly effective, especially when combined with individualized feedback.

Ultimately, multi-pronged strategies including feedback and social influences are most effective.

Long term change is only achieved when social norms change. People must be convinced of the effectiveness and value of their efforts.

Slide 10: Findings
From our experience with existing approaches and our review of the literature, we’ve arrived at several key findings informing our own approach to the problem.

Combine different motivational techniques in order to amplify the effects of each.

Address multiple perspectives on energy conservation by catering to various motivations and mental models. For example, one person in the household might think in economic terms and be best motivated by financial units, while another may be concerned with emissions and be motivated by information expressed in environmental units such as kg/CO2.

Provide multiple methods of access to the same information: make it accessible rather than demanding the resident refer to a centralized source.

Support in-the-moment, situated decision-making as well as long-term planning.

Define baselines of energy consumption appropriate to individual circumstances. Turns out that in general the wealthiest individuals use the most resources, and can also make the greatest conservation gains due to this high baseline. This is a unique situation that won’t necessarily translate to a low-income household or a shared living space.

Combine feedback with opportunities for change. For example, situate information about heating costs alongside a thermostat control.

Slide 11: An Ecosystem Approach
We’ve applied these insights and arrived at an ecosystem approach to the problem.

Seeks to provide information where and when it’s needed, integrated into the home as a holistic system.

Slide 12: ALIS
This approach is exemplified by our implementation of ALIS: the Aware Living Interface System.

In North House and West House, we developed prototypes across a web application, social networking tools, a mobile application developed for the iPhone, touch-based controls, and ambient and informative art displays.

Provides tools that are useful for residents both at home and away, when going about daily activities, and when actively thinking about resource use in the home.

I wish I had time to talk about each piece in detail, but I’m going to quickly explain our approach to ambient and artistic feedback.

Slide 13: Ambient & Artistic Feedback
These kinds of tools:

  • Display information that is important but not critical
  • Can move from the periphery to the focus of attention and back again
  • Provide subtle changes to reflect updates
  • Are aesthetically pleasing and environmentally appropriate

Slide 14: Ambient Canvas
One major prototype we developed in this space is called the Ambient Canvas.

It’s a light display embedded in the kitchen backslash of West House made up of LED strings mounted behind acrylic and glass.

It provides feedback on residential resource use by displaying relative levels of use.

Slide 15: Water Use
It can be configured to provide feedback on water use.

As the residents go about their daily activities like washing dishes, taking a shower, running the washing machine and so on, the Canvas will fill up to display their use as a relative value compared to their average use.

During the Olympics when we were exhibiting West House, we saw that visitors were really drawn to this display. They’d come into the kitchen and say “Oooohh, that’s pretty!” Then they’d pause and say, “…What is it?”

So after hearing this a few dozen times we knew we were on to something, but that conveying information in a non-traditional format takes careful consideration.

Our demonstration had the Canvas fully lit. In our model, this would indicate a high level of resource use – something to be avoided! However, visitors loved the look of it when fully lit, so perhaps our model is backward. Perhaps visually active states should be reserved for low levels of use in order to encourage conservation.

Slide 16: Next Steps
In my research, I’m trying to answer these and a wide variety of other questions about the development of these systems.

Our next steps are the refinement of concepts and further development of prototypes, experimental evaluation of ambient and artistic feedback, and eventually long-term studies in West House.