Hardware is where constraints become visible.
Eight years across five thermostat products and connected sensing devices—from fixed-segment displays to touchscreens, serving contractors, DIY homeowners, and hotel guests. Each had its own way of failing, yet all of them had to feel like one family. What hardware taught me is that coherence is never styling—it is judgment, one call at a time.
Patents
US 12,608,066
Low Power Detection
Awarded · 2026
Patent pending
Segment display character set
Photo 1 of 5
The screen technology changed, the experience had to change with it.
Touch 1 mirrored the fixed-segment Sensi Smart next to it in the product line — a reasonable first step, but it left most of what a touchscreen could do on the table. For Touch 2 I kept Touch 1's menu structure intact (no reason to make users relearn navigation) and rebuilt the visual language, interaction model, and interface architecture underneath, after a teardown of Touch 1's strengths and weaknesses, a competitive scan, and interviews with utility partners, contractors, and homeowners, each of whom hit different problems.
I came in during industrial design evaluation, where bezel width, thickness, and screen size all move the interface's reach and density, so I sat between ID, engineering, and UI — turning form-factor constraints into interface decisions and interface needs into hardware requirements. Touch 2 shipped at 0.77", the thinnest smart thermostat in its category at launch, and the interaction model later fed into consulting work with Trane's own touchscreen team.
A fixed contextual action button, same spot on every screen, always set to the most important next step. Users learn it once, and it holds everywhere — nobody notices consistency like that until it breaks.
Remote sensors and the Equipment Interface Module were nowhere on the roadmap when I laid out the framework, but I left room for them. When they arrived later, they slotted in without a rewrite — I treated the architecture as a design decision, not an engineering afterthought.
The design problem wasn't digital, it was physical.
The Equipment Interface Module bridges thermostats to HVAC equipment on installs where the existing wiring can't carry what the system needs. Competitors covered that range with three separate modules; the Sensi EIM covers it with one. The default assumption was that setup happened at the equipment — meaning trips back and forth between thermostat and unit, often through a cramped mechanical space, time a contractor can't bill for. So I moved configuration to the thermostat: the contractor pairs the two devices, sets equipment type and location once, and the EIM configures itself. No return trips.
Two distributor groups brought up the pairing process unprompted in post-launch training sessions, and the flow generated zero iteration requests after launch — contractors are vocal when something doesn't work, so silence carries information too.
The launch webinar for the EIM pairing flow drew more than 180 contractors — direct exposure to exactly the audience whose install time the redesign was trying to save.
Thirty-two segments, three buttons, no room to hide.
Sensi Lite is the most constrained product I've worked on: thirty-two display segments, three buttons (up, down, menu/action), and a brief for a complete thermostat experience that still had to feel premium. With three inputs, hierarchy is too expensive, so navigation stays flat and cyclical — the menu/action button does different jobs depending on how it's pressed: a normal press advances, a long press opens homeowner settings, a second long press reaches contractor configuration, keeping settings that can break an HVAC system within reach for installers and out of the way for everyone else. With flat navigation, order does the work a menu hierarchy usually would, so I sequenced settings by dependency (heat pump setup leads straight into reversing-valve direction), most common configuration first.
Sensi Lite also runs on power stolen from the HVAC system, so when power drops too low for the display to run, the device still has to communicate state. I helped work out a prioritized shutdown sequence and a notification flow that pushes device state to the mobile app, so the phone stands in for the display when the thermostat can't — the low-power detection method is the subject of US Patent 12,608,066. Two of my calls lost stakeholder review, and post-launch data proved both right, which taught me to put reasoning on the record before losing an argument.
In usability testing, 23% of users stalled at the three-dot menu button because it didn't read as pressable, and auto mode made 30% hesitate — a sign the conditional sequencing was pulling its weight. Icon testing ran five variants and settled on a square inside a circle, lifted from the media stop symbol; that mark carried into later products.
I sat in on physical installs, watching people tuck wires, snap the unit onto its baseplate, and fight with the clips, then turned what I saw into hardware requirements — clip resistance, baseplate fit, where the controls sat. On a product this small, UI and physical design stop being separate problems.
Designing for guests who will never be onboarded.
Verdant serves hospitality: guests who touch the thermostat once, speak different languages, and will never read a help screen — the interface has to land immediately or it's failed. For the European Line Voltage model I dropped text entirely and built the interface out of icons, since no caption could bail out a language gap. The character set itself needed to read from across a hotel room, not up close, so I studied existing segment forms, drew ideal numerals in Lato, then overlaid every digit on transparency paper to find which segments always fire together and could be merged — trading some accuracy on rare letterforms (W, Q) for a smaller, more legible set.
The finalists went through A/B testing on PCB prototypes in a room sized like a hotel room (15–20 feet of viewing distance), with participants who'd never seen the product, varying lighting and text at random. I led the design of the character set, and Canada has granted the design patent, with a US application still pending. Examiners cleared it against prior art that included a competitor's own thermostat display patent, confirming the design was differentiated within the category and not just novel in the abstract.
The wireless sensor protocol built for Verdant's hardware later became Sensi's room-sensor network — not because the products share an ecosystem, but because the underlying problem was identical and the work already existed. When the same problem shows up twice, reaching for what already works beats walling it off by product line.
