Fusion PCB Rendering

Final revision 2 with JTAG debugger

Mapping an FFT to a rainbow gradient

Fusion rendering of PCB + enclosure

Front view of product

Schematic designed using Eagle

Assigning "weights" to frequency ranges and pushing them out from the center

Assigning color length from frequency range amplitudes

Revision one (SGLT5000 + ESP-WROVER-B)

Routing in Eagle

Example of react-native color palette control

Bluetooth audio-reactive LED bar

Palette LED Controller

trade hounds cover

Trade Hounds is a react-native app designed as a swiss army knife for people in the trades. &quot;Hounds&quot; are able to build an online portfolio of their best work, browse work from Hounds around them, see job offers in their area, anonymously compare their pay, chat with users, and search for used tools and supply houses around them. The app has over 250,000 users, and has been live on the AppStore and Google play since early 2018.Features<ul><li>Mapbox-powered maps: two of the main application features are map-focused and use mapbox + a tiling server to serve data at various layers. This data includes anonymous pay data submitted by users that&#39;s bucketed into regions, and POI information on supply houses and used tools.</li><li>In-app campaigns powered by backend-served modals: pop up modals are configurable JSON objects that the front end picks up on in network response interceptors. This gives the backend the flexibility to serve a modal in response to any request.</li><li>Social feed: users are able to upload text, photo, and video posts to the application, which can sort the posts by all users or only followed users. By leveraging image caching, view recycling, and efficient rendering, the feed is kept responsive.</li><li>Low memory footprint: the native app and react-native javascript app are audited monthly for memory usage with the goal of staying under 650mb, despite being a media and map-heavy application. Some ways in which this is done:<ul><li>AWS Media Services are used in a lambda when uploading content to compress and resize content.</li><li>Re-rendering is kept at a minimum and only for necessary UI changes.</li><li>UI and business logic that can be separated into a native module are kept away from the javascript thread.</li><li>In-app videos are mounted and re-used as a singleton view, with only it&#39;s source being swapped out. So there is very minimal deallocation and reallocation of memory-heavy videos.</li><li>View hierarchy depth is kept at a minimum, and libraries like react-native-screens helps with this.</li><li>Views that can be torn down on de-focus are removed from memory.</li><li>Persisted application state is kept at a bare minimum so as to keep app launch fast.</li><li>All UI animations and gestures occur off of the javascript thread using react-native-reanimated and react-native-gesture-handler.</li><li>React-native and important libraries are kept up to date.</li></ul></li><li>In-app chat: using sendbird and a local caching solution, users are able to send messages to each other reliably and in a variety of network conditions. File and text messages will attempt to re-send if they previously failed to send, and will send in the background when possible.</li></ul>

Masonry-grid style discovery feed for browsing user-uploaded content

In-app campaigns for winning prizes based on completion of certain MAU-driving activities

Mapbox-powered pay transparency feature where users can compare their pay to similar users around them

Social feed containing posts, questions, and polls from users and trades you follow

Chat screen for text and media messages between users

Digital portfolio and social space for tradespeople

Trade Hounds

Wide angle photo of ceiling

Screenshot of MadMapper configuration

Video of lava animation

Rack unit containing a Mac Mini and Chauvet Net X II

iOS app with color HSL color palette control

iOS app with cue control

Marquee-style madmapper lighting system

Friki Tiki, NYC

recover cover photo

Recover Athletics is an iOS/Android react-native app designed to help aid in the prevention and treatment of running-related injuries. Users are asked to give post-run qualitative data on soreness which powers custom pre and rehab routines. Over time, users can track their progress and tweak the in-app routines that work best for them. The app is integrated with <a href=https://www.strava.com/ target="_blank">
 Strava
 </a> and <a href=https://www.garmin.com/ target="_blank">
 Garmin
 </a>, so logged activities can be overlaid with user-generated injury reports.Routines were filmed in-house and reviewed by some of the top runners and rehab professionals in the Boston area. Recover Athletics was puchased by Strava in 2022 and is now part of their formal product suite.

Main activity screen where users can follow recovery-focused workouts

Home screen where users can select pre-made and custom workouts for their specific recovery needs

Swipe-up bottom sheet view with information on current routine

Post-routine summary screen where users can track their progress over time and try similar routines

Daily logging screen where users can enter qualitative data about their injuries

Mobile app for running recovery aid

Recover Athletics (Strava)

good karma cover

Good Karma DAO&#39;s mission is to set a new standard for how the music industry operates and how artists are treated. Holding their membership NFT grants a vote within their DAO (one vote per wallet). It offers the chance to choose the artists that GKR signs, and have a say in all of the day-to-day operations of the label. The NFT will also grant token gated access to special events virtual and in-person, early exposure to signed artists&#39; music, and free entrance to GKR concerts.
Ultimately, the goal of the mobile app is to be the one-stop shop for all GKR related NFT and account management activities. Today, the app serves content from GKR&#39;s <a href=https://www.contentful.com/ target="_blank">
 Contentful
 </a> account, which includes blog posts, news announcements, and events. Every time new content is posted on Contentful, a webhook is sent to a <a href=https://firebase.google.com/ target="_blank">
 Firebase
 </a> function that triggers a rich push notification to all subscribed users. Users have fine-grained control over which types of content they receive push notifications for. Firebase is also used to manage mobile authentication and user settings.
The article detail screen is especially prominent in this app, with rich-text html display capabilities and modal overlays that let the reader explore artists without ever leaving the post. Clicking an artist name, for example, will pop a bottom sheet that displays information about that artist, images, social links, and streaming links.

Animations for launching article detail screen

Event detail screen with calendar integration

Home screen integrations and launching event detail screen

Article detail screen

Onboarding experience

Home screen with content fed from Contentful

Push notification settings

Mobile app for $KARMA token holders

Good Karma

Full drum kit running a rainbow animation

Designed for band <a href=https://bandboston.com/ target="_blank">
 Boston's
 </a> 2017 HyperSpace Tour, this system triggers <a href=https://cdn-shop.adafruit.com/datasheets/WS2812B.pdf target="_blank">
 WS2812b
 </a> leds on every drum hit using piezo sensors. The system is low-profile, waterproof, and stress-tested for road conditions. An on-board, high-contrast OLED screen allows settings to be customized on the fly, and noise thresholds adjusted and saved for drummers.

Custom PCB for Teensy 3.2 connections to WS2812b leds

Video of drums being used on tour

Inside of computer housing

Reactive LED drum kit for band Boston

HyperSpace

larry's flight cover photo

Larry&#39;s Flight is a WiFi-enabled Art-Net LED cloud system for <a href=https://www.aeronautbrewing.com/ target="_blank">
 Aeronaut Brewery
 </a> in Everett, MA. The title of the piece sticks with the Brewery&#39;s cloud/flying theme, originally based off the <a href=https://en.wikipedia.org/wiki/Lawnchair_Larry_flight target="_blank">
 Lawnchair Larry Flight
 </a> (that inspired the movie Up).
Both clouds are powered by an ARM Cortex-M7 <a href=https://www.pjrc.com/store/teensy41.html target="_blank">
 Teensy 4.1
 </a> which outputs an <a href=https://art-net.org.uk/ target="_blank">
 Art-Net
 </a> (sACN) ethernet signal, but also accepts an external signal with a higher priority. This allows any lighting designer capable of outputting sACN to override the system and sync the cloud lighting with music and stage lighting. Once they disconnect, the computer reverts back to the default animations (custom <a href=https://fastled.io/ target="_blank">
 FastLED
 </a> routines) which are controllable by the brewery staff via a react-native app that communicates over WiFi with the system using a <a href=https://docs.particle.io/photon/ target="_blank">
 Particle Photon
 </a>.
The clouds themselves are mainly chicken wire and polyfill held together with spray glue and a small amount of fishing line. Inside each cloud is a PVC pipe with 300 and 450 <a href=https://cdn-shop.adafruit.com/datasheets/WS2812B.pdf/ target="_blank">
 WS2812b
 </a> leds , respectively. The data signal from the main computer is converted to the required SPI signal using an <a href=https://dmxking.com/led-pixel-control/ledmx4-pro target="_blank">
 LeDMX4 PRO
 </a>, and carried to the lights using Cat6 cable.

Finished cloud #1 (of 2)

Workshop during lighting system build

Rack unit with Particle Photon and Teensy 4.1

Final cloud placement in Aeronaut Brewery

Rack system, front view

Example of FastLED animations

Workshop setup with two clouds

React-native app with animation and color palette control

WiFi-enabled Art-Net LED cloud system

Larry's Flight

View from outside the structure

&#39;INSOUND&#39; is an immersive light and sound installation where visitors can experience what it might be like to be inside of a piece of music. By re-mapping the audio spectrum onto various color gradients, users experience a synesthetic connection between light and sound.
The technology to power the sound to light conversion is powered by the <a href=https://www.mix-sig.com/index.php/msgeq7-seven-band-graphic-equalizer-display-filter target="_blank">
 MSGEQ7
 </a> frequency band analysis chip. This is hooked up to a ARM Cortex-M7 <a href=https://www.pjrc.com/store/teensy41.html target="_blank">
 Teensy 4.1
 </a> running animations using the <a href=https://fastled.io/ target="_blank">
 FastLED
 </a> library, which ultimately outputs a DMX signal to 7 high-power DMX bar lights.
The physical material for the structure was provided by <a href=https://molodesign.com/ target="_blank">
 molo designs
 </a>, and is a soft polyester material with a rigid steel frame.

Drew André is a Boston-bred full stack software engineer and lighting designer currently working at LAZ Parking in San Diego.