PROJECT OVERVIEW
Project type: Meter User Experience research for Boond Solar India
Team: UX Researcher/Designer; Boond Solar CTO, Engineers and Coordinators; Advisors at Aalto University Design and Business Schools
Date: Sept 2017- June 2018, Part-time MA Design thesis project
Skills: plan, coordinate and conduct ethnographic design research, UX Design
Methods: contextual inquiry, physical prototyping, affinity mapping, visual design
Tools: Adobe CC, video, photography, transcriptions, Excel spreadsheets
THE CHALLENGE
Although Boond’s solar-powered electricity service is a life-changer for rural Indian villagers, consumption is much lower than expected. Why? To find some answers, I was tasked as a Research Assistant for Aalto New Global to investigate customers’ consumption habits and interactions with their electricity meters, and to propose design improvements to the meter and service experience.
I had limited prior knowledge about solar-powered electricity or the rural Indian context, but that didn’t deter me. The opportunity to do field research that could positively impact sustainable energy uptake motivated me to take on this challenging role.
A customer shows us her electricity meter installation and the devices she charges.
FIELD RESEARCH GOALS & METHODOLOGY
Goals
The three key research questions we wanted to answer by going into the field were:
- How do customers interact with and understand their electricity meters?
- What are customer’s daily household interactions with energy sources?
- How does the meter’s smart pricing influence behaviour?
Methodology
I started with secondary research to understand the broader context
- At the beginning of the project I knew very little about solar-powered electricity. I love getting immersed in new topics, so I dug in by reading articles and books about social issues relating to rural electrification and development. I also familiarised myself with historical and contemporary electricity generation and metering and discussed the local situation and service challenges with the Boond CTO.
Next, I planned and conducted field-based design research in India using contextual interviews and design props to prompt discussion and evaluate the meter’s usability.
In the field I collaborated with customers, a small field team of language/cultural interpreters and Boond’s local team.
FIELD-BASED DESIGN RESEARCH
Recruitment criteria and process
We used snowball recruitment for the contextual interviews. Life in the villages is fluid and unpredictable, so a flexible recruitment method to match the context was needed.
- The local team contacted the village chief in advance to set up the first customer interview of the day. While the interview was underway the local Boond team found the next participant through their village network.
- Usually we’d start by interviewing an individual. Often a family member (or more) would join in once the interview was rolling.
The language and cultural interpreters conduct an interview with a father and son, while the female family members look on. During this interview I observed and took photos and asked questions to build on my observations and follow-up on previous comments.
Contextual Interviews
I conducted contextual interviews with 24 participants in 19 homes in four villages over six days. I asked the participant to show me their electricity meter, lighting setup, appliances and any other energy sources they used. Participants also showed me how charge their phone(s) and described their daily phone use.
- The local field team and I collaborated on operational tasks including documenting the research, problem-solving and helping to shape the fieldwork approach on a daily basis.
- Once participants agreed to take part in the study, they actively collaborated in the research by showing me their homes, describing their daily lives, interacting with the prototypes and helping with recruitment.
We discovered that school-age family members often read the meters for their parents or relatives. This young boy steps up on a sack of rice to read the meter.
Design Props
One of the big challenges I faced in the field research was evaluating if, and how, textual non-literacy [1] posed a barrier to understanding the meters. To better understand how textual non-literacy played into metering, I made two design props, using early design ideas I got from the contextual interviews.
- I quickly redesigned the meter interface and put it on a cardboard mockup of a re-designed meter.
- The prop was a tool to explore, with sensitivity to the stigma of non-/low-literacy, how participants respond to the use of Hindi language and icons instead of the English text on the meters.
- I also used the prop to find out how participants think of electricity use in terms of time and spending.
I made the design prop made with a printout of a new UI design pasted onto a cereal box and an iPhone with an Hindi interface mocked-up for the display.
Using the prop to facilitate discussion with the participant who shows and tells us how he understands the new meter design.
I made a second design prop to get feedback from participants on graphics ideas gleaned from the first round of interviews.
Insights from the Design Props
- Participants who can read, read Hindi not English, the meter language (children often read the meters, and are learning to read Hindi in school).
- Participants recognise and understand the word "balance" and their balance amounts (previously learned by using prepaid mobile phones).
- The so-called “universal” icons in my design were not recognised by the participants.
- For competing energy sources (kerosene, batteries) participants think of spending in months, versus spending in days and weeks for prepaid electricity.
Reflection on the Props method
- Although I used what I thought were global tech icons, my visual language was very different from that of the participants. Given more time, it would be better to co-create the visuals with the users.
- When I showed the meter Engineer my design, he told me that the language on future meters could be customized to other regions of India. Given the number of languages in India, using text in the local language (supported by visuals) would be a more compelling solution to pursue.
ANALYSIS & INSIGHTS
Data synthesis methods included affinity mapping and visual analysis of field research photos. I organised the participant's energy use into a matrix to understand and communicate the allocation of energy source in relation to low and high demand of task and perceived low to high cost of the energy source.
Analysis & Synthesis
After a day in the field I made daily field memos. Back at my desk, I reviewed my field notes, coded the interview transcripts and did visual analysis on media and recorded data on a spreadsheet. For synthesis I made affinity diagrams, an energy use matrix, and discussed the findings with Boond’s CTO.
Key Insights
In answer to the research questions, I learned critical things about the users in the following three themes:
1. How customers interact with and understand their electricity meters:
Children read the meters for their parents and help them manage spending.
Many users can’t see the meter because it is installed too high up on the wall, they climb up on a table or sack of rice to take meter readings.
Users check the meter for their account balance and to estimate the next recharge.
Except for the account balance, the technical data shown on the meter display is irrelevant to the user.
Interview excerpts:
“Boond recommended to put [the meter] up this high, to keep it out of reach of the kids.”
“I do not look for watts, only rupees. I am not literate that much to check other things”
“I check the meter 3-4 times a day and give updates to my Mom.”
A customer’s young grandson learned to read the meter from his Uncle; he stands on his toes to check the meter balance a couple of times a day for his grandmother. A sketch from my field notes captures an early observation that the meters were installed too high on the wall and should be lowered to make them accessible to the diversity of people that use them.
2. Customer’s daily household interactions with energy sources:
- Users very cleverly and frugally use a mix of energy sources for specific tasks, based on perceived cost:
- kerosene for soft outdoor lighting at night;
- auto batteries for powering appliances;
- their own solar home systems for phone charging;
- Boond’s bright lighting is used for studying and household security. Users are unclear about their monthly spending on the service. This makes it difficult to compare the service costs to their energy mix that they know well and still use.
Interview excerpts:
[on the patio at night] ... “we switch on Boond, do some work, then switch it off and burn a kerosene lamp.”
“We charge the phone on our own Solar Home System because it doesn't cost anything. We don’t pay anything additional.”
The mix of energy sources that we saw being used in participants’ homes included an auto battery (charged by a diesel engine at a local depot) used for charging mobile phones and kerosene burned in lamps at night.
3. How the meter’s smart pricing influences behaviour:
In a nutshell, it doesn’t. Users weren’t aware of smart pricing. This is partly because the meters don’t display price differences clearly and partly due to shortcomings in meter instructions and/or service onboarding.
Interview excerpt:
“If we will use one small bulb so it will charge one rupee for an hour. And a tube light will charge for double the amount.”
[The actual cost is roughly half this amount.]
Additional discoveries: thinking beyond the meter
Although the service delivery was modelled on very successful prepaid mobile phone plans, the electricity meter just isn’t capable of supporting the rich communication that smart metering requires (tight frugal constraints limit the meter’s functionality).
This capability gap could be filled by transitioning some aspects of metering to a mobile phone. Most families had at least one feature phone in the household. With growing mobile network access and smartphone ownership in rural India, phone metering could solve many of the problems that users, and the company, face. My thesis included an SMS service design proposal (not covered in this case study) to address this opportunity.
DELIVERABLES
The core problem customers encountered is difficulty monitoring their electricity consumption. This is because the meters weren’t designed for the wide range of users in the households: multi-generational rural-Indian families, with varying levels of textual- and digital-literacy.
Under frugal constraints (limiting design changes to the meter interface only) I gave recommendations for three key upgrades, and illustrated them in a mock-up of a new meter user interface (shown in the before/after image above).
1. To overcome textual low-literacy and language barriers:
- Change the meter language to Hindi and redesign the icons. LCD displays have several limitations that hinder rich textual treatments. Ideally, the LCD display should be in Hindi, but, this requires a larger LCD display, which the current meter design won’t accommodate.
- After exploring several options, my solution proposed using English on the display with a translation printed above. An added benefit of this approach is the meter can be adapted to local languages.
The LCD display in English with a Hindi translation printed above considers flexibility in regional languages as Boond’s business expands across India.
2. To help users understand electricity costs, consumption spending and change behaviour:
Use everyday language on the price indicator lights/control labels
Print a price chart on the front of the meter (a previously underutilised space).
Eliminate the technical data and fit customer’s consumption data onto two rotating screens.
Rationalise and align the text and numbers for easier to reading.
Screen One (above) shows the account balance and accumulated monthly consumption. Seeing spending over the longer term helps users compare their spending with other energy sources they use (kerosene, auto battery).
Screen Two (below) shows real-time consumption in rupees and in watts per hour. This shows transparency in pricing (to build trust) and tells users their current spending to run appliances (educate on costs).
The LCD display in English with a Hindi translation printed above considers flexibility in regional languages as Boond’s business expands across India.
3. To make the meter physically accessible to users of all abilities and heights, lower the meter installation.
- Position the meter at eye level where adults and school-age children can view the display.
- Add a security latch and position the meter out of reach of toddlers for added protection.
The existing meter (before) and my retouched version of the photo (after). I retouched the photo to show first-hand how lowering the meter could dramatically improve users’ access to the display.
IMPACT
At the product level, Ideas from the redesigned meter interface were adopted by the Boond Engineer and included in the next generation meter interface, which was printed in the Hindi language.
At an organisational level, the research demonstrated a human centered design process and gave the Boond team new customer insights.
Project feedback
From Master's thesis Evaluator Professor Jack Whalen:
“The execution of this field research is first rate. All the evidence from the thesis indicates the interviews were carried out with considerable care and sensitivity (close attention to context and each subject's specific concerns). The observational work […] was also quite carefully done and yielded many insights. This was no small feat, given that Simmonds is not Indian and does not speak Hindi, having to rely on a translator and work in an environment very different indeed from that with which she is most familiar.”
CHALLENGES
As a non-native of India, understanding context, facing the language barriers, encountering lags and interruptions in translation during interviews made it challenging to clearly grasp some participants responses. I missed some opportunities to learn more. A daily debriefing and involving the field team in synthesis would have helped to overcome some of these shortcomings.
The meter redesign process would have benefited from greater user participation. Next time, were I given the budget, I would explore and evaluate prototypes with participants. I would also collaborate with a local design partner to facilitate cross-cultural expertise.
A spread from the book I designed for my Master’s Thesis, Power to the People: Designing a Better Prepaid Solar Electricity Service for Rural Indian Villages.
The research uncovered insights about the wider service, and I proposed further design improvements that I would be happy to discuss further. The full report may be found in my Master’s thesis, Power to the People: Designing a Better Prepaid Solar Electricity Service for Rural Indian Villages, available for download on Aalto University’s archive.
Field research photos by Lindsay Simmonds and Iba Marwein.
Notes
- I refer to Jan Chipchase’s discussion on literacy: “The term 'textually non-literate' reflects that there are many ways to define literacy. For example, task-literacy can be the ability to complete a particular task, computer-literacy the ability to make basic use of a computer.” I acknowledge that, like Chipchase, “non-literacy is not caused by lack of ability but rather by lack of opportunities for learning”.
Say hello!
In need of a UX Researcher? Comments or questions? I'd be happy to hear from you!