This category is a bit of a catch-all and contains both historic projects as well as some fun projects. As far as the historic projects are concerned, those date back to before becoming independent in 2012 but deserve being mentioned here (see list below) because they represented great engineering challenges. In contrast, the fun projects are projects we took on because we love solving engineering problems (instead of doing cross-word puzzles or Sudoku...) and, best of all, were able to help somebody out.
Solving multi-disciplinary engineering problems, both large and small.
Sometimes we do things simply because we are interested in the subject. Like that latest engine concept... does it really work? Or a product idea intrigues us. Or we decide to create a device because there is nothing on the market that works. And sometimes, we lend a helping hand to somebody pursuing a - in our judgement - worthy cause.
We use our experience to solve some things that from an economic perspective would not be worthwhile. But you know what? It can make all the difference for those receiving the help. We are thankful that this consulting business has given us the freedom to do just that.
Sample Project: Point Clouds from Stereoscopic Imaging
Background: 3-dimensional point clouds can be very desirable for not only identifying objects within the field of view (FOV) but also for quantifying their position and orientation in 3-d space. The two primary methods employed today for producing 3-d point clouds are time-of-flight (ToF) cameras, including Lidar, and stereoscopic imaging based on at least two cameras viewing an object from slightly different directions.
Features: This software and associated tools were written as a feasibility study for creating point clouds. It utilizes the well-known OpenCV library for camera calibration and point cloud generation. In comparison to commercial offerings (Intel RealSense D435) and competing ToF cameras (Meere MDC1200), the results were competitive but showed artifacts typical of stereoscopic imaging due to shadowing and texture effects (see dropouts in picture below). Various improvements are possible but have not been pursued in this feasibility study.
Application: Categorization of objects based on color and shape
List of Select Other Projects
Historic project: Fully automated image processing
and pattern recognition for biotech applications and
hardware that enabled rapid processing of thousands
of samples per hour.
This was a cross-disciplinary project and we were involved
in all aspects of software, hardware, applications and
CHT Model Validation: Building of prototype enclosure
and instrumentation to validate key aspects of electronics
cooling during operation.
Measured transient temperatures, flowrate and back pressure
under various operating conditions and compared with results
from CHT analysis obtaining excellent correlation.
Ultrasonic Curb Sensor: Engineered sensor solution to
reliably detect curbs (common in parking lots) or severe
transitions in slope (common between driveway and road)
to prevent potential damage to front spoiler/splitter of
Competing technologies use infrared light that can be
defeated by curbs with reflective paint in parking lots.