AI/ML Tech
Android & WearOS | iOS & watchOS & visionOS | Robotics & IoT (Rust)
Machine Learning on Mobile
HummingBird Fuel
This AI-powered mobile app utilizes machine learning to analyze health data and recommend personalized microdosing schedules for medication. This has the potential to optimize medication effectiveness while minimizing side effects. Our initial focus is on using caffeine for targeted energy boosts.
Website: https://www.hummingbirdfuel.com/
Photodo
This AI-powered to-do app lives conveniently on your phone. Simply capture a photo, record a voice memo, or take a note, and Photodo's intelligent assistant swings into action. It identifies the task, sets deadlines, locates nearby assistance, and even tracks your budget. It's like having a personal AI assistant in your pocket. (Android)
DrunkWatch
This innovative system leverages your smartwatch to assess your intoxication level and determine your driving suitability. It analyzes your walking patterns, writing style (when entering a code), and bio-sensor data to make this crucial call. This showcases the combined power of the iWatch's 4-core Neural Engine, accelerometer, and bio-sensors. (watchOS / iWatch10)
Research: https://zoewave.medium.com/life-saving-watchos-ml-app-c9fde03e79cf
Website: https://www.zoewear.com/
RxDigita
Streamlining medication management, RxDigita allows you to add meds by voice or photo (the app reads the label!), set reminders, and keep track of everything in a clear and organized list view. (Android)
Website: https://www.rxdigita.com/
AI Appliance Nutritionist
Abstract: The rise of smart appliances presents a unique opportunity to revolutionize personal nutrition. This article explores the concept of an AI/ML-powered appliance nutritionist, a system that integrates with smart refrigerators and ovens to guide users towards healthier eating habits. We discuss the potential functionalities of this system, including personalized meal planning, real-time recipe suggestions, and automated cooking based on dietary needs, available ingredients and current health data.
Article: https://zoewave.medium.com/ai-ml-appliance-nutritionist-8d9f8a1c84d5
Gemini AI
Google Fit / Health Connect
Android
Self-Driving LEGO® CarÂ
This project explores the exciting concept of creating a self-driving LEGO® car with your smartphone acting as the brains of the operation.
Design Docs:
https://zoewave.medium.com/self-driving-lego-car-concept-c3a50e72cccc
AshBike
This eco-friendly e-bike service leverages your fitness data (via Apple Health and Google Fit integration). Its ML capabilities help plan routes based on your calendar schedule and reward you with NFTs for choosing sustainable transportation. (Android & iOS)
Website: https:www.ashbike.com
AI/ML Security Micro-Drone
Firefly Autonomous Home Security Micro-Drone
Micro-drone capable of autonomously surveying a home while the residents are away. This drone will capture images, upload them to YouTube, and notify the homeowner of any significant changes detected within the environment.
We will delve into the components, software, and functionalities required for this project, emphasizing safety and responsible use.
https://zoewave.medium.com/autonomous-home-security-micro-drone-df989ea02075
Small Businesses with AI/ML
Implementing AI/ML to generate foot traffic and enhance customer understanding marks a transformative step for local mom-and-pop shops. By harnessing user interaction data, demographic insights, and behavioral patterns, we can create sophisticated models that personalize customer experiences and provide actionable insights for merchants.Â
~~~
⬅️ This article emphasizes the critical role of foot traffic for the survival and success of small businesses. It highlights that small businesses make up 99.9% of all U.S. businesses but often struggle to survive, with many closing within a few years. To help these businesses, the article suggests using AI/ML to increase foot traffic, proposing a mobile app that shows local shops offering free samples. This approach aims to drive customer visits, provide real-time feedback, and enable A/B testing for brick-and-mortar stores, ultimately boosting sales and customer loyalty.
https://zoewave.medium.com/saving-small-businesses-with-ai-ml-2dfed17248a7
AI/ ML Soda Fountain
This concept reimagines movie theater soda through AI/ML. Users create custom drinks on a smartphone app, with AI suggesting flavors based on movie genre (think citrusy thrillers or berry-licious comedies!). This personalization enhances the moviegoing experience and caters to individual preferences.
~~~
Interacting with the AI Soda Fountain will put the movie goer into the mode of movie with a unique AIGen Interaction (voice, sound & images):
Scary drinks with mysterious, frightening
Comedy drinks with silly, funnyÂ
RomCom drinks with loving, fun
Action drinks with fast paced, fast movingÂ
https://zoewave.medium.com/re-imagine-ai-ml-soda-fountain-8555143036b2
AI/ML Creativity
Facial Aware Ads (FAPA)
We're excited to present our research on Facial Aware Personalized Advertising (FAPA), a novel approach that personalizes in-mall (offline) advertising without compromising user privacy.
Imagine a future where digital billboards adapt to each viewer. As a shopper walks by, the display analyzes their facial features (not for identification!), emotions, and even clothing style to deliver the perfect ad in real-time. For example, a surprised woman carrying a gym bag might see a fitness tracker ad, while a stressed-looking man in a suit could be shown a relaxing massage promotion.
FAPA leverages cutting-edge AI/ML technologies like convolutional neural networks to create relevant and engaging ad experiences. Crucially, our system prioritizes user privacy by anonymizing all data before analysis.
https://zoewave.medium.com/facial-aware-personalized-advertising-fapa-4b88ea668cb8
Jag.U.R™ Utility Robot
Jag.U.R™ : Rust--AI/ML assistive/service helper RoboDog.
JourneyAgileGuiding - Utility U: Robotic R:Â
Pronounce: Jag.U.R (dʒæg.jʊ.ar) ™ = "Jag.You.Are"
We can look forward to a future where intelligent robots seamlessly integrate into our lives, empowering us and shaping a brighter tomorrow. Imagine a world where robots like Jag.U.R. become trusted companions, empowering us to live richer and more fulfilling lives.
https://zoewave.medium.com/jag-u-r-utility-robot-d6045e215b20
AI Generated Mobile App
Gen AI (ChatGPT) made this mobile app with very little help.
Article: https://medium.com/@zoewave/ai-fixed-coding-9076db6a4a19
Demystifying Mobile Machine Learning
A Multi-Tool Approach
Android's & iOS ML Arsenal: Choosing the Right Tool
When it comes to Android & iOS , there's no one-size-fits-all solution for ML.Â
Here's a breakdown of the key options and their ideal use cases
Android
ML KitÂ
Perfect for quick integration. It offers user-friendly APIs and pre-built models for common tasks like face detection or text recognition. However, customization is limited.
Firebase ML
Firebase ML
Seamlessly integrates with the Firebase ecosystem, providing pre-built models for specific tasks like image labeling or language translation. Similar to ML Kit, customization options are restricted.
TensorFlow (TF) Lite
TensorFlow (TF) Lite
If high performance and complete control are your priorities, TF Lite is the way to go. It allows you to leverage custom models, but requires a solid understanding of machine learning concepts.
MediaPipe (New)
MediaPipe (New)
This newcomer simplifies complex machine learning pipelines by offering a versatile and easy-to-use framework built on top of TF Lite. For developers seeking a low-code or no-code alternative to TF Lite, MediaPipe is a great option.
Gemini Nano (New)
Gemini Nano (New)
Privacy takes center stage with Gemini Nano. This secure execution environment boasts a chat-like interface for interacting with models. However, current limitations include flexibility and device compatibility (currently only available on Pixel 8 Pro).
AI Core (New)
AI Core (New)
This hardware acceleration component complements Gemini Nano on compatible devices, further optimizing its performance.
iOS
iOS
Core ML and Create ML - A Powerful Duo
Apple's mobile ecosystem offers Core ML and Create ML for streamlined ML development on iOS:
Core ML
Core ML
The core (pun intended) of on-device machine learning in iOS apps. It acts as a bridge, allowing your app to interact with various machine learning models in a unified way. These models can be used for tasks like image classification or making predictions. Core ML empowers you to run models directly on the device, enhancing speed and privacy compared to cloud-based solutions.
Create ML
Create ML
This user-friendly tool streamlines the creation of custom machine learning models. Imagine it as a drag-and-drop interface where you provide your training data, and Create ML handles the model creation process. It's ideal for beginners or scenarios where complex models aren't required.
Detailed Breakdown of Autonomous Home Security Micro-Drone
This document outlines the design and functionalities of an autonomous home security micro-drone system, leveraging advancements in miniaturization, sensor technology, and software.
Components
Micro-Drone Hardware
Flight Components
Micro-drone Frame: A lightweight and durable frame, typically constructed from carbon fiber or high-strength plastics, is crucial for maneuverability and weight limitations.
Brushless Motors: High-efficiency brushless motors provide sufficient power for desired flight times and agile movements.
Battery: A lightweight, high-capacity Lithium Polymer battery offers extended flight times (around 10 minutes).
Power Management Unit (PMU): The PMU regulates power supply to the flight controller and other electronics, ensuring stable voltage and preventing damage.
Battery Monitoring IC: A dedicated integrated circuit offers real-time monitoring of battery voltage and current, essential for estimating flight time and triggering low-battery warnings for safe returns. (Not required for smartphone flight controllers)
Micro-Drone Surveillance System
Camera: A small, high-resolution camera with Wi-Fi capabilities is necessary for image capture.
Wi-Fi Module: Enables real-time image transmission from the camera to the control system.
Communication Ports: These ports allow the flight controller to communicate with other components and receive control signals. Common ports include:
Serial Ports (UART): Used for communication with telemetry radios, GPS modules (if applicable), and external configuration devices.
I2C and SPI: These internal communication buses allow the flight controller to interact with peripherals like sensors and actuators.
Distance Sensor: An ultrasonic sensor or LiDAR system helps the drone navigate obstacles and maintain safe flight paths.
Inertial Measurement Unit (IMU): The IMU is a collection of sensors that measure the drone's orientation, acceleration, and rotation. A 6-axis IMU combines a gyroscope, accelerometer, and magnetometer for comprehensive data.
Gyroscope: Measures the drone's rate of rotation around its roll, pitch, and yaw axes.
Accelerometer: Measures the drone's acceleration in all directions.
Magnetometer: Measures the Earth's magnetic field, aiding in compass functionality and orientation. (Smartphones typically have powerful IMUs)
Barometer: This sensor measures the surrounding air pressure, which helps the flight controller maintain altitude and improve flight stability, especially during indoor navigation.
Optional: Onboard SD Card Slot: An SD card slot allows for data logging, helpful for troubleshooting flight issues or analyzing performance.
Microcontroller Unit (MCU): This is the central processing unit of the flight controller. A mid-range to high-performance 32-bit MCU is recommended for autonomous tasks. Some popular options include:
G4: Offers a balance of processing power and efficiency.
F7: Enables faster processing speeds.
H7: Leverages even faster dual cores, ideal for complex flight algorithms and sensor fusion.
Smartphone Flight Controller
A smartphone can be repurposed as the flight controller due to its processing power, sensor data, and communication capabilities. This eliminates the need for a separate flight controller.
Open-source autopilot software options include PX4 Autopilot and AndroCopter, which allow for using a smartphone to control the drone.
Software and Functionality
Flight Planning Software: Software is required to program the drone's autonomous flight path within the home environment. This path should leverage AI/ML algorithms to ensure complete coverage and collision avoidance.
Image Recognition Software: Software with object recognition capabilities will analyze captured images and compare them to baseline images taken when the home is unoccupied. Significant changes, such as the presence of unauthorized objects, can trigger alerts.
Automatic Image Upload: The captured images will be automatically uploaded to a private YouTube channel accessible only to the homeowner.
Base Station
Charging, Alarm/Alert & Data
An alert system will notify the homeowner via email or text message in case of significant changes detected in the images or if the drone goes missing from the charging station.
A dedicated charging station with a docking mechanism ensures efficient charging and secure storage of the drone between flights. The station can be equipped with sensors to detect missing drones.
Rust Security Server
A Rust server running on the base station provides functionalities like:
Secure storage for captured images.
Secure communication with the drone.
User authentication for access control.
Building and Programming
Building the micro-drone involves careful assembly of hardware components, followed by configuration of the flight controller