Case Studies

FingerSense

 

Qeexo AutoML enabled FingerSense to scale to over 200 million devices as the top value-add software.


 

FingerSense is the world’s first software-only solution that can distinguish between touchscreen inputs such as fingertip, knuckle, nail, and stylus. It is powered by Qeexo’s lightweight machine learning technology and leverages data from existing sensors on the device.

 

Huawei, the world’s third largest smartphone manufacturer, uses Qeexo FingerSense technology (rebranding it as “Knuckle Sense”). With FingerSense, Huawei smartphone users gain access to many powerful tools. For example, users can use a knuckle to knock twice on the display to take a screenshot, or to draw the letter “M” using a knuckle to open a user-defined music app. FingerSense was rated top value-added software in Huawei Device Co—36.6% of Huawei users frequently use FingerSense*.

 

The demand for FingerSense grew rapidly after launch, and Qeexo’s ability to meet the demand from customers became limited by the traditional machine learning development process. Development times were lengthy, visibility into the data was poor, and the workflow required constant oversight and error checking. In order to meet the demand of nearly 100 commercial projects per year, Qeexo developed AutoML.

Case Study

Without Qeexo AutoML

  • Labor-Intensive: Each project is an individual machine learning problem that requires tuning from machine learning engineers.
  • Time-Consuming: From data visualization and preprocessing to machine learning library deployment, it requires an average of 20 hours for one machine learning engineer to implement one project.
  • Difficult to Scale: FingerSense was implemented in over 90 different commercial projects in 2016 alone. Without Qeexo AutoML, this would not have been possible.

With Qeexo AutoML

  • Automatically preprocess data, generate visualizations and performance report, and build machine learning libraries using multiple algorithms to compare metrics, all without needing help from machine learning engineers.
  • AutoML allows non-machine-learning-experts to build the same production-ready models and reduces model development time by 95%.
  • Machine learning engineers were freed up for more critical and complex R&D tasks, such as performance improvements and optimization research. With this reclaimed time, they were able to enhance FingerSense’s overall performance by 15%.
  • Qeexo AutoML enabled the deployment of FingerSense on 200+ million devices worldwide in just four years.
  • FingerSense was rated top value-added software in Huawei Device Co—36.6% of Huawei users frequently use FingerSense*.
Case Study

Magic Wand

 

Easily enhance product intelligence with Qeexo AutoML; rapidly iterate through prototypes to reach commercial product.


 

One toy manufacturing company in Taiwan (under NDA; denoted “Company” below) wanted to produce a toy magic wand. With the use of accelerometer and gyroscope sensors, the wand would play different sounds and display different colors when performing various gestures in the air. Unfortunately, the accuracy for identifying the different gestures was unsatisfactory and the product could not be commercialized prior to Qeexo AutoML.

Without Qeexo AutoML

  • Lack of knowledge and expertise to process sensor signals and utilize sensor data.
  • Lack of budget and resources to hire qualified machine learning engineers.

With Qeexo AutoML

  • Qeexo AutoML automatically extracts features and develops optimal models to allow Company to select the best-possible machine learning model for their use case, thereby improving the accuracy and overall usability of the toy.
  • Qeexo AutoML allows Company to take Magic Wand to production without having to hire a team of machine learning engineers.
  • Machine learning models generated by Qeexo AutoML are designed to have low latency and a small footprint and work extremely well in highly-constrained environments like Company’s Magic Wand, which has low memory space, processing power, and battery life.
  • Company is planning to produce a series of “smart toys” using Qeexo AutoML.

Quality Inspection in Automotive Factory

 

Qeexo AutoML builds machine-vision-based quality inspection to automatically vet for defective parts in smart factory.


 

KwangSung Corporation is an automotive parts manufacturer and Tier 1 supplier for Hyundai Motor Company and Kia Motors with its operations in Korea, America, China and India. It wanted a better way to detect defects in “quarter garnish” parts, including scratches, gas marks, short shots, and black spots. Before AutoML, KwangSung employed workers to visually inspect parts with their bare eyes.

Without Qeexo AutoML

  • Manual labor is expensive and error-prone. Employee turn-over rate is high.
  • Alternatives were extremely expensive and needed modifications to current manufacturing setup.

With Qeexo AutoML

  • Qeexo AutoML does not require expensive infrastructure and equipment, and can be integrated with a small fraction of the cost and effort.
  • Qeexo AutoML can be used by existing KwangSung employees, without complicated training and ramp-up time to operate.
  • Machine learning models generated by Qeexo AutoML run on the Edge and do not require internet connectivity, so the data stays on-premise.
  • KwangSung is rolling out Qeexo AutoML to its other factory sites worldwide after initial evaluation in Alabama.
Case Study
How can AutoML help your Organization?

Please include country code if outside of the United States