Toward the Design of an Energy Consumption Feedback System

Abstract

In this article, we examine two novel issues in user requirements analysis for a feedback system on household energy consumption: (a) microgeneration and (b) the “Must, Should, Could, Won’t have” (MoSCoW) method. We report on a qualitative user requirements analysis using the MoSCoW method for the prioritization of user requirements. Fifteen interviews resulted in three user groups that were abstracted into personas with different user requirements. Finally, we show that changes in cutoff values resulted in MoSCoW prioritization changes for a maximum of only 5.9%. These are promising results for the robustness of the MoSCoW method.

Keywords

green energy automation design persona sustainability user groups energy consumption qualitative research system design user requirements analysis human-automation interaction

Qualitative research methods identified three personas – the innovator environmentalist, the technology user, and the saver – and design features to satisfy a range of needs.

To attain the goal of sustainable energy consumption, consumers seek A-label domestic appliances, green energy suppliers, or electric transportation. In addition, they perform many behavioral energy-saving actions, such as isolating their homes, turning off unnecessary lights, and turning down central heating by a few degrees.

Yet, for many years, actual energy consumption could not be perceived directly; it was abstract, untouchable, and invisible to the consumer. These days, energy consumption has become more visible, and the important role of feedback on energy-consuming behavior has been well recognized. However, technological developments prompt new ways for energy consumption and feedback. These technological developments place energy consumption feedback in the broader framework of human factors/ergonomics (HF/E) in automation design (Lee & Seppelt, 2012). The design entails not only visual feedback displays but also human–automation interaction.

In this article, we examine two novel issues that have arisen partly because of technological developments:

How should a feedback system for microgeneration be designed?

How robust is the MoSCoW (“Must, Should, Could, Won’t have”) method for the prioritization of user requirements?

Existing Literature

In the following sections, we discuss the role of feedback on energy consumption and the prioritization of user requirements.

Feedback

Results from the Fischer (2008) review, based on 21 original papers on the effects of feedback on energy consumption, clearly showed that feedback stimulates energy savings. Most reported savings are between 5% and 12%. The pitfall is that inadequate feedback may undermine human–automation interaction (Lee & Seppelt, 2012). This area is where HF/E principles and user-centered research and design come into play. As HF/E specialists know, a successful feedback system begins with an understanding of the needs and requirements of the users, as specified in the ISO 134-7 standard (International Organization for Standardization, 1999; Maguire & Bevan, 2002). “A user requirement comes from a user or other type of stakeholder and expresses a property of the domain or business process that the introduction of a new system will bring about” (Maiden, 2008, p. 90).

Including user requirements analysis early in the design phase may prevent user dissatisfaction with the system and high cost afterward because of adjustments of superfluous systems. User requirements for a feedback system on household energy consumption are well established (e.g., Darby, 2006; Fischer, 2008).

MoSCoW method

The review papers mentioned earlier lack a method for the systematic prioritization of requirements. In addition, Maguire and Bevan (2002) and Spath, Hermann, Peissner, and Sproll (2012) acknowledged the importance of prioritization to ensure that effort is put into the most important aspects of the system, yet they proposed no actual method for doing so. Traditional research used more quantitative approaches, such as simply counting (Roberts, Humphries, & Hyldon, 2004) or computing percentages of those in favor of a certain requirement (Sernhed, Pyrko, & Abaravicius, 2003). The MosCoW method requires quantitative counting and the computing of percentages as well, yet the advantage is that it combines these with qualitative value assessments of users’ comments concerning user requirements.

Novel Issues in Energy Feedback Requirements

Our user requirements study had a strong qualitative research focus in which the experience of the user was central. We conducted individual interviews with 15 participants from the LochemEnergie project (see next section) between the ages of 40 and 75 (M = 54.3, SD = 8.2). The interviews were transcribed and analyzed following the qualitative data analysis from Baarda, De Goede, and Teunissen (2005; see Supplemental Materials A, available online at http://erg.sagepub.com/supplemental for analysis of the interview).

There is a new development toward microgeneration, whereby households become generators as well as consumers of green energy. This decentralization of control over power grids is one of the technological developments that may help overcome the pitfalls of managing complex networks of automation (Lee & Seppelt, 2012). LochemEnergie is such a project that aims to accelerate the transition from fossil and nuclear energy resources to renewable, green energy sources by microgeneration. The project is one of the earliest and most innovative sustainable energy initiatives in the Netherlands.

Because of the innovative nature of the LochemEnergie project, its members should be categorized as early adopters (Rogers, 1995) of sustainable energy consumption. This project combines green energy and a smart grid, also known as an intelligent grid (Farhangi, 2010). The smart grid is necessary because LochemEnergie deals with unstable energy supplies and demands that are too complex for the conventional grid. Energy should be consumed directly because it cannot be stored for later consumption.

Together with the smart grid, LochemEnergie aims to motivate its members to change their behavior and actively match Lochem’s locally generated energy with local consumption. The results discussed in this article portray the user requirements regarding this novel issue of microgeneration.

A relatively new data analysis method is the MoSCoW prioritization method. According to Tudor and Walter (2006), this method is well suited for hierarchical classification of user requirements in a quantitative manner (Brennan, 2009). Requirements are systematically assigned to one of the following four categories: “must have,” “should have,” “could have,” and “won’t have.”

Unfortunately, there is little literature on the robustness of the MoSCoW method and how it should be used. The main issue concerns the cutoff values that are used to determine MoSCoW categories and whether this determination of requirement categories is a reliable and valid method. Because the method is relatively new, there are no guidelines for defining cutoff values in the existing literature. However, different cutoff values would probably lead to different prioritization of the requirements. Therefore, we aim to take the first step in establishing the robustness of this method.

How Should the Feedback System for Microgeneration Be Designed?

After we analyzed and obtained the user requirements from the transcribed interviews, the question remained: Which user requirements are most vital to the success of the feedback system and which are less important?

Prioritization

We used the MoSCoW method to prioritize the user requirements according to their relative importance into one of the four categories: “must haves,” “should haves,” “could haves,” or “won’t haves.” The method consisted of three steps (see Supplemental Materials C, available online, for a more extensive description of these steps):

The number of participants with a positive attitude toward a requirement was established. A positive attitude was considered an attitude that was in favor of a certain requirement.

The strength of individual attitudes was established by qualitatively assessing the strength value of the attitude.

The combined MoSCoW scores were calculated for each requirement by computing the average of first two steps.

User requirements

The user requirements were divided into information and medium/design requirements. We report only the “must haves” and “should haves” that resulted from the analysis because they are most vital to the success of the feedback system.

Table 1 provides an overview of the information requirements, and Table 2 provides an overview of the medium/interface requirements. The tables also describe practical implications for design.

Table 1.

“Must Have” (boldface type) and “Should Have” (regular type) Information Requirements and Implications for Design

User Requirement	Implications for Design
Generation (kWh)	Users want to know how much energy they generate every day, preferably linked to weather conditions. The user wants to see generation displayed in kWh. A combination of kWh and euros is also a good option.
Matching generation/consumption	Users want to know if their energy generation matches their consumption. Feedback on an individual scale is most important, but feedback on the Lochem Scale is also an option.
Energy exchange	Users want feedback about the energy exchange between houses, LochemEnergie, and supplier. Information should clearly show consumers what kind of energy they use and whether it comes from LochemEnergie or another supplier. Information should not be too specific.
Direct	Users want direct feedback that links behavior at a specific time and context with the energy generation/consumption at that specific time.
Historical comparisons	Users want to see trends in their own generation/consumption behavior over different periods. These can be years, seasons, months, days, weeks, or adjustable periods.
Electric car	Users want to know when their electric car is used to balance the power grid. This information should be available in both the car and the house.
Looking back	Users want their data to be saved so they can look back for up to 4 years at least.
Frequency on demand	Users want their direct feedback to be on demand. The indirect feedback can have a standard frequency, such as once a week, month, or quarter. Users also want to be able to adjust these periods.
Normative comparisons	Users want to benchmark their own energy generation and consumption with that of other households. Comparisons should be reliable and with comparable houses, and users want to know how comparable they are.
Indirect	Next to on-demand feedback, users also want indirect feedback. The indirect feedback should have more of an overview function, showing the bigger picture of LochemEnergie.
Advice/tips	Users want to receive advice on how to better match generation and consumption.

Table 2.

“Must Have” (boldface type) and “Should Have” (regular type) Medium/Interface Requirements and Implications for Design

User Requirement	Implications for Design
Application on multiple devices	Users want their feedback through an application that can be accessed by means of at least tablets, computers, and smartphones. Users also want to be able to print their feedback and an option to receive an information overview by e-mail. Elderly users want text to be in an appropriate letter size that enables them to easily read the feedback.
Zooming function	Users want their feedback to be presented hierarchically and with a zooming function. Zooming needs to be possible from the general to the more specific breakdown in terms of time (years, months, weeks, days, hours, minutes), and from graphs to more detailed information in numbers.
Display influence	The user wants to be able to influence the display. There should be a standard format, but users want to choose which items they would like to add or delete from their “front page.” Also offer the opportunity to choose to display feedback in graphics, text, or both. For consumers who do not appreciate display influence, different user styles can be an option.
Automatic matching	Users want their generation and consumption to be matching automatically to some extent. They do want to have total control over which appliances they allow to be turned on automatically. The user also wants an overruling function and option to set a maximum starting time. Be cautious with implementation and do not make it a standard function, because some people feel strongly negative toward this function.
Bar graph for comparisons	Users want bar graphs for comparisons.
In one glance	Users want a FrontPage style that is very clear and understandable and that allows consumers to interpret feedback in one glance.
Practical/easy to use	Users want an application that is practical and easy to use. Usability should be tested before implementation.
System simulation	Users want the system to simulate the effects on consumption and generation of particular actions. Simulation should be reliable, intelligent, and tailored to the specific household. Simulation can be about new behaviors or (generation) appliances.
Alert	Users want to be alerted when their preset boundaries are crossed. Boundaries also can be based on consumers’ own data history. The alert should be in the form of a color or light signal on tablet, computer, or smartphone. The alert should not be too intrusive.
Graphical	Users want graphical feedback.
Colors	Users would like to see clear colors that already have meaning, such as red and green. Another option is colors that go from light to dark.
Line graph	Users prefer line graphs that show matching of energy generation and consumption.

User profiles

The notion of different user profiles emerged during the interview analysis. Some user characteristics and motivations seemed to cluster into more general user profiles with different user requirements. Automation designed to perform in a manner consistent with operators’ preferences, expectations, and individual mental models improves human–automation interaction (Lee & Seppelt, 2012). Therefore, we chose to consider these (collateral) results as highly important and to investigate this additional notion by using the persona technique from the human–computer interaction domain (Cooper, Reimann, & Cronin, 2007).

The persona technique

A persona is an abstraction of a group of real users who share common needs and characteristics (Pruitt & Adlin, 2006). The persona is represented through a fictional individual, who in turn represents a group of real users with similar characteristics and goals (Pruitt & Adlin, 2006; Turner & Turner, 2011).

The main benefit of personas is that they help focus software analysis and design on the characteristics and goals of the product’s end users (Cooper & Reimann, 2003). They are also a good method for enhancing engagement and reality, which makes them more memorable to designers (Grudin & Pruitt, 2002). We used six of the seven steps from the design process for personas of both Cooper and Reimann (2003) and Castro, Acuña, and Juristo Juzgado (2008). The steps were as follows:

Identify behavioral variables that serve as the basis for different personas.

Map the scores of participants on each behavioral variable on a scale, which will result in a clustering of participants into more general personas.

Identify significant behavioral patterns for each persona.

Identify relevant characteristics and goals for each persona.

Check whether previously identified aspects are fully defined in the personas.

Create a third-person narrative about a typical day in the life of each persona and complement this narrative with a fictional name and picture to complete the persona. (See Supplemental Materials A, available online, for a more detailed description.)

Following these steps, we could distinguish three user groups within the user sample: the innovator environmentalist, the technology user, and the saver. Figures 1, 2, and 3 portray excerpts of these personas that resulted from the analysis (the full personas can be found in Supplemental Materials B, available online).

Figure 1.

Excerpt of the innovator environmentalist persona.

Figure 2.

Excerpt of the technology user persona.

Figure 3.

Excerpt of the saver persona.

Differences in user requirements

The results show that different user requirements are appropriate for different personas. Table 3 provides an overview of the differences in user requirements.

Table 3.

Differences in User Requirements Between Personas and Implications for Design

Persona	Requirement	Implications for Design
The innovator environmentalist	Accessible	The innovator environmentalist values accessible, practical, and easy-to-use feedback. Conduct usability/accessibility testing before implementation. Although mostly mentioned by the innovator environmentalist, all users will benefit from such feedback.
The technology user	Normative comparisons	The technology user wants normative comparisons. Technology should ascertain that comparisons are indeed comparable.
	System intervention	The technology user wants a system intervention function. This user wants total control over which appliances will be automatically turned off. Also provide an on/off option. Caution is advised with implementing this function. It should not be a standard function, as other personas feel negatively toward system intervention.
	Smart grid	The technology user wants a system that is intelligent, is tailored to specific households, and makes optimal use of the smart grid.
The saver	Indirect	The saver values indirect feedback. Such feedback should have more of an overview function, showing the bigger picture of LochemEnergie. This should be an option only, as not all personas are equally interested in this function.
	Energy-saving game	The saver persona wants an energy-saving game. The saver wants the game to be supportive, educational, and fun for all age groups. This should be an option only, as not all personas are equally interested in this function.
	System simulation	System simulation is valued by the saver. It should be intelligent, reliable, and tailored to the specific household. Simulation can be about behaviors or new (generation) appliances. This should be an option only, as not all personas are equally interested in this function.

Only differences larger than successive categories are shown because of the robustness of the MoSCoW method, as shown in the next section.

How Robust is the MoSCoW Method for the Prioritization of User Requirements?

The robustness of the MoSCoW method was examined by making four changes in the cutoff values. The resulting changes in prioritization of all information requirements (n = 82) were obtained. The results show that changes in cutoff values result in category changes only between successive categories. For example, “should haves” can become “must haves,” but “won’t haves” cannot become “should haves” or “must haves.” Small, but also larger, changes in cutoff values result in combined MoSCoW score changes for a maximum of only 5.9% of the content requirements. Table 4 shows the cutoff changes and their resulting changes in MoSCoW prioritization. More detailed changes in the two noncombined MoSCoW prioritizations can be found in Supplemental Materials C, available online.

Table 4.

Changes in Cutoff Values and Their Resulting Changes in MoSCoW Scores

Type of Change	Percentage of Changes in Combined MoSCoW Scores	Changes
Shift of 0.1 in strength of quotation	0%	—
Shift of 1 in number of participants	5.9%	W→C 1C→W 1S→C 1M→S 2
Combination of shift in 0.1 in strength of quotations and shift of 1 in N of participants	3.7%	W→C 1C→W 1S→C 1
Large uneven shift in strength of quotation	1.2%	S→C 1

Note. MoSCoW = Must, Should, Could, Won’t have. Arrows indicate change (e.g., W→C = “won’t haves” become “could haves”).

Conclusions and Recommendations

The goal of the study outlined in this article was to find out how a feedback system for microgeneration should be designed and to investigate the robustness of the MoSCoW method. The most important conclusions and recommendations are discussed next.

Microgeneration

Early-adopter households found the following information most important (i.e., “must haves”) for a feedback system for microgeneration: historical comparisons, direct feedback, generation, matching of generation and consumption, electric car, looking back, and on-demand feedback. “Must haves” for the medium and interface were application, zooming function, display influence, and automatic matching. The feedback should be presented graphically; use bar graphs for comparisons, line graphs, and colors. The microgeneration feedback system should entail as much “must haves” and “should haves” identified in this study as possible.

Can the results from this early-adopter sample be generalized toward the larger population? It should be noted that these findings confirmed user requirements, elicited among all adopter categories, from existing energy consumption literature and complemented them with microgeneration user requirements. Therefore, we assume that user requirements elicited from innovator and early-adopter households can be generalized to some extent to the general population. Authors of further research should examine more systematically whether this presumption holds true.

Different types of feedback for different users

The investigation into the different user groups that resulted from the data led to the conclusion that different types of feedback are appropriate for different types of users. This conclusion can be explained by the findings that individuals may have different levels of norm activation and motivations (Matthies, 2005). In addition, users may be in a different stage of behavioral change, as proposed by He, Greenberg, and Huang (2010). These differences are especially evident in functions concerning automation and learning.

We recommend that when designing a feedback system, designers should not use a one-size-fits-all solution for different user groups. When dealing with early adopters and under pressure of limited resources, we recommend that designers target the innovator environmentalists first, followed by the technology users and then the savers. This recommendation is based on the relative dominance of these various types of users within the sample, with the highest prevalence shown by the innovator environmentalists (n = 7) based on their strong convictions and motivations toward sustainable energy consumption. The detailed personas and their different preferences should serve as a basis for further research and prototyping of a feedback system for sustainable energy consumption and microgeneration.

With regard to which persona should have priority in design, one should examine whether the distribution of the general population among personas is comparable to that of the present early-adopter sample, as this distribution could also have important consequences for determining priorities.

The MoSCoW method

Changes in cutoff values of this relatively new method resulted in category changes only between successive categories. Our investigation into the robustness of the method showed that the method is robust against changes in cutoff scores. These results are promising for future use of this method. However, a more systematic investigation is necessary in order to make more reliable claims regarding the robustness of this method.

Based on this first test, the MoSCoW method seems to be a robust and promising technique for relatively quick and hierarchical classification of user requirements. The method would greatly benefit from research that establishes clear guidelines for its optimal use.

Overall, we recommend continuing work on this line of qualitative user requirements research. A great advantage of this study is that it translates a large amount of rich qualitative research data into more practical and workable data for system designers. Practical and workable data are especially important when working on sustainable energy consumption, because professionals with engineering and social science backgrounds have to work together to find the most sustainable and optimal solutions. There is a need for a bridge between these two groups of researchers. The MoSCoW method, combined with detailed and vivid personas, can form this bridge and thereby decrease the gap between the system designer and the actual needs of the end user.

Footnotes

Suzanne Beltman received her MSc in human factors and media psychology from the University of Twente. Her research looks at user-centered and energy-related work in human factors.

Suzanne M. Vosslamber is a teacher of human factors and engineering psychology at the University of Twente, the Netherlands. Her interests include energy-related work in human factors. She has an MSc in human factors and media psychology from the University of Twente.

Albert Molderink received his PhD degree in computer science from the University of Twente. Nowadays he is working for Ipsum Energy as researcher and as part-time assistant professor at the University of Twente. He is working in a research group that investigates the possibilities of increasing energy efficiency using embedded control, mainly via optimization and control algorithms.

Matthijs L. Noordzij is an assistant professor of human factors and engineering psychology at the University of Twente, the Netherlands. His research looks at user-centered design of both products and processes related to (energy) technology. In addition, he is involved in several projects in which ambulatory physiological measurements are integrated in care and work settings.

References

Baarda

De Goede

Teunissen

(2005). Basisboek kwalitatief onderzoek [Basic book qualitative research]. Groningen, Netherlands: Stenfert Kroese.

Brennan

(2009). A guide to the business analysis body of knowledge (Babok Guide). Toronto, Canada: International Institute of Business Analysis.

Castro

J. W.

Acuña

S. T.

Juristo Juzgado

(2008). Enriching requirements analysis with the personas technique. In Proceedings of the First Workshop on the interplay Between Usability Evaluation and Software Development (pp. 13–18). Pisa, Italy: CEUR.

Cooper

Reimann

(2003). About face 2.0: The essentials of interaction design. Indianapolis, IN: Wiley.

Cooper

Reimann

Cronin

(2007). About face 3.0: The essentials of interaction design. Indianapolis, IN: Wiley.

Darby

(2006). The effectiveness of feedback on energy consumption: A review for DEFRA of the literature on metering, billing and direct displays. Oxford, UK: Environmental Change Institute, University of Oxford.

Fischer

(2008). Feedback on household electricity consumption: A tool for saving energy? Energy Efficiency, 1, 79–104. doi:10.1007/s12053-0089009-7

H. A.

Greenberg

Huang

E. M.

(2010). One size does not fit all: Applying the transtheoretical model to energy feedback technology design. In Proceedings of the 28th International Conference on Human Factors in Computing Systems (pp. 927–936). New York, NY: ACM. doi:10.1145/1753326.1753464

Farhangi

(2010). The path of the smart grid. IEEE Power and Energy Magazine, 8(1), 18–28. doi:10.1109/MPE.2009.934876

10.

Grudin

Pruitt

(2002). Personas, participatory design and product development: An infrastructure for engagement. In Proceedings of the 7th Biennial Participatory Design Conference (pp. 144−152). Malmø, Sweden.

11.

International Organization for Standardization. (1999). ISO 13407: Human-centred design processes for interactive systems. Geneva, Switzerland: Author.

12.

Lee

J. D.

Seppelt

B. D.

(2012). Human factors and ergonomics in automation design. In Salvendy

(Ed.), Handbook of human factors and ergonomics (4th ed., pp. 1615–1642). New York, NY: Wiley.

13.

Maguire

Bevan

(2002). User requirements analysis. International Federation for Information Processing, 99, 133–148.

14.

Maiden

(2008). User requirements and system requirements. IEEE Software, 25(2), 90–91.

15.

Matthies

(2005). Wie können PsychologInnen ihr Wissen besser an die PraktikerIn bringen? Vorschlag eines neuen, integrativen Einflussschemas umweltgerechten Alltagshandelns [How can psychologists improve their outreach to practitioners? A suggestion for a new, integrative model of environmentally sound everyday practice]. Umweltpsychologie, 9(1), 62–81.

16.

Pruitt

Adlin

(2006). The persona lifecycle: Keeping people in mind throughout product design. San Francisco, CA: Morgan Kaufmann.

17.

Roberts

Humphries

Hyldon

(2004). Consumer preferences for improving energy consumption feedback. Report for Ofgem. Bristol, UK: Centre for Sustainable Energy.

18.

Rogers

E. M.

(1995). Diffusion of innovations (4th ed.). New York, NY: Free Press.

19.

Sernhed

Pyrko

Abaravicius

(2003). Bill me this way! Customer preferences regarding electricity bills in Sweden. In Proceedings of the 2003 Summer Study of the European Council for an Energy Efficient Economy (pp. 1147−1150). Stockholm, Sweden: European Council for an Energy Efficient Economy.

20.

Spath

Hermann

Peissner

Sproll

(2012). User requirements collection and analysis. In Salvendy

(Ed.), Handbook of human factors and ergonomics (4th ed., pp. 1615–1642). New York, NY: Wiley.

21.

Tudor

Walter

G. A.

(2006). Using an agile approach in a large, traditional organization. In Proceedings of AGILE 2006 Conference (pp. 367−373). Washington, DC: IEEE Computer Society.

22.

Turner

(2011). Is stereotyping inevitable when designing with personas? Design Studies, 32(1), 30–44.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.96 MB