Helia
Redesigning a smart home energy dashboard
Helia is a concept redesign of a smart-home solar energy dashboard, built for everyday homeowners rather than engineers. The original monitor buried people in raw numbers, so we rebuilt it into a calm, plain-language home screen that shows, at a glance, how your panels are doing, what you are saving, and the impact you are making. All wrapped in a warm brand made for high-end homes.
Helia
Redesigning a smart home energy dashboard
Helia is a concept redesign of a smart-home solar energy dashboard, built for everyday homeowners rather than engineers. The original monitor buried people in raw numbers, so we rebuilt it into a calm, plain-language home screen that shows, at a glance, how your panels are doing, what you are saving, and the impact you are making. All wrapped in a warm brand made for high-end homes.
Product
B2C
Design Systems
Role
Product & Motion Designer
Timeline
Feb - Mar 2026
Credits
Framer is a no-code tool for building and publishing responsive websites—perfect for anyone creating modern, high-performance pages without coding.
Problem
How do you make a solar dashboard a normal person can actually read?
Six categories of information competed for attention at once, all in raw numbers, in a layout built for data instead of people. For a homeowner who is not an engineer, a line like “Performance ratio: 92%, down 4%” means nothing, even though they are the one who paid for the panels. The data was all there, but the meaning was not. Underneath it sat three questions the product never actually answered: Is my system working? Am I really saving money? Is this good for the planet?
Problem
How do you make a solar dashboard a normal person can actually read?
Six categories of information competed for attention at once, all in raw numbers, in a layout built for data instead of people. For a homeowner who is not an engineer, a line like “Performance ratio: 92%, down 4%” means nothing, even though they are the one who paid for the panels. The data was all there, but the meaning was not. Underneath it sat three questions the product never actually answered: Is my system working? Am I really saving money? Is this good for the planet?
Research
We benchmarked the whole category, then turned every metric into a question.
Brand. We studied energy and solar brands, including Tesla Energy, Tibber, Sense, SolarEdge, Enphase, Span, and Iberdrola, looking at their typography, logos, color, photography, and illustration. The pattern was clear. Most leaned cold and technical. Warm tones read as energy and sun, green read as eco, and cold blues read as pioneering or technological. Almost nobody was designing warmth for a premium home, and that became our opening.
Information. In parallel, we reframed each of the six required data categories as a question a homeowner would actually ask: what is happening right now, are my panels at full capacity, what should I expect tomorrow, how much am I really saving, how much of what I used was free, and what is my real impact. Once each block was a question, the hierarchy and the plain-language headlines followed.
We kept testing informal, with classmates and supervisors rather than a formal study, but it changed real decisions.
Research
We benchmarked the whole category, then turned every metric into a question.
Brand. We studied energy and solar brands, including Tesla Energy, Tibber, Sense, SolarEdge, Enphase, Span, and Iberdrola, looking at their typography, logos, color, photography, and illustration. The pattern was clear. Most leaned cold and technical. Warm tones read as energy and sun, green read as eco, and cold blues read as pioneering or technological. Almost nobody was designing warmth for a premium home, and that became our opening.
Information. In parallel, we reframed each of the six required data categories as a question a homeowner would actually ask: what is happening right now, are my panels at full capacity, what should I expect tomorrow, how much am I really saving, how much of what I used was free, and what is my real impact. Once each block was a question, the hierarchy and the plain-language headlines followed.
We kept testing informal, with classmates and supervisors rather than a formal study, but it changed real decisions.
Iteration
The layout got simpler with each round of feedback.
The scroll took two reviews to get right. We started with a dense view showing every widget at once, then a version split into three fixed screens. Both times, feedback pushed us simpler, until a single continuous scroll won.
Iteration
The layout got simpler with each round of feedback.
The scroll took two reviews to get right. We started with a dense view showing every widget at once, then a version split into three fixed screens. Both times, feedback pushed us simpler, until a single continuous scroll won.
Scrollable, what we chose
Compact view, what we cut
Decisions
The home screen reads top to bottom by priority.
Lead with language, not numbers. Every card opens with a sentence, like “You’re storing energy” or “Your panels are maxing out,” with the figures kept as backup. An ambient AI layer, not a chatbot, does this on its own: it reads the live data, writes the headline, and flags what needs attention.
The home screen reads top to bottom by priority: a real-time hero, then system health, tomorrow’s forecast, savings, the daily solar-versus-grid balance, and environmental impact. A modular grid lets every widget resize and reflow. We ordered each widget by how often someone checks each thing, most-glanced at the top, deep detail below or behind a “See more.” The tradeoff is that detailed performance sits low, which is wrong for an installer but right for the everyday homeowner we were designing for, and they were always the priority.
Decisions
The home screen reads top to bottom by priority.
Lead with language, not numbers. Every card opens with a sentence, like “You’re storing energy” or “Your panels are maxing out,” with the figures kept as backup. An ambient AI layer, not a chatbot, does this on its own: it reads the live data, writes the headline, and flags what needs attention.
The home screen reads top to bottom by priority: a real-time hero, then system health, tomorrow’s forecast, savings, the daily solar-versus-grid balance, and environmental impact. A modular grid lets every widget resize and reflow. We ordered each widget by how often someone checks each thing, most-glanced at the top, deep detail below or behind a “See more.” The tradeoff is that detailed performance sits low, which is wrong for an installer but right for the everyday homeowner we were designing for, and they were always the priority.
Connected devices, my idea, sit in a persistent panel one tap away, rather than buried in settings, following a pattern people already know from smart-home apps.
Connected devices, my idea, sit in a persistent panel one tap away, rather than buried in settings, following a pattern people already know from smart-home apps.
A warm brand, not a techy one. Backed by the benchmarking and aimed at the real buyer, high-end families. One small touch: the environmental card counts flowers planted, tying back to the flower-shaped logo.
To keep three people moving in the same direction, we set up a lightweight design system early. Tokens handled colour and spacing, and every component had its states built in, charging, alerts, low battery, so the dashboard stayed consistent as it grew.
A warm brand, not a techy one. Backed by the benchmarking and aimed at the real buyer, high-end families. One small touch: the environmental card counts flowers planted, tying back to the flower-shaped logo.
To keep three people moving in the same direction, we set up a lightweight design system early. Tokens handled colour and spacing, and every component had its states built in, charging, alerts, low battery, so the dashboard stayed consistent as it grew.
A lightweight system of tokens, components, and states
Reflection
What I would test with another month.
This was a concept, shaped by benchmarking and critique rather than live users, so what I take from it is mostly a list of things worth proving. Whether plain-language headlines genuinely help people understand their energy faster. How much people trust an AI reading of their own data, and what it would take to earn it. Whether the frequency order holds once real households use it, or whether everyone checks something different. None of it needs a redesign. It needs real users.
Reflection
What I would test with another month.
This was a concept, shaped by benchmarking and critique rather than live users, so what I take from it is mostly a list of things worth proving. Whether plain-language headlines genuinely help people understand their energy faster. How much people trust an AI reading of their own data, and what it would take to earn it. Whether the frequency order holds once real households use it, or whether everyone checks something different. None of it needs a redesign. It needs real users.
The making of, powered by matcha, mostly
