A Guided Leap over FPGA Knowledge Gaps Drives Innovation in Aerospace

Posted on September 12, 2019 by Matthew Russell

The Wright Brothers may have known relatively little about the physics of flight when compared to modern avionics engineers, but that obviously didn’t stop them.

Just 36 years after the Wrights devised a multi-winged marvel in their bicycle shop, the world’s first jet aircraft took flight. Jet airliners were ferrying passengers through the sky 13 years later, and when the Boeing 707 began taxiing runways by the end of the 1950s, the avionics industry began to soar.

The Wright brothers’ third test glider is piloted by Wilbur Wright, with Orville Wright is at left, and the brothers’ friend Dan Tate at right.

Great advances in aerospace are still being made today, though they may not be as cosmetic. Embedded technology has driven much of the aerospace industry’s innovation in the last few decades because of the space constraints inherent in even the most massive aircraft. Tightly arranged aircraft parts are never far from a centralized embedded system, either, and sensors in each can easily be networked together to gather important performance and environmental data to trigger the appropriate responses.

In the form of multiple high-speed cameras attached an airplane, you have the makings of a “smarter” aircraft that can log image data alongside the other measurements. Technology like this can aid in disaster response, provide highly-detailed maps, and assist airborne law enforcement in major interdictions. Without a means to accurately time stamp the data and a processing system that can match it up in order to inform safety-critical decisions, however, it’s just expensive aerial photography equipment.

The right IP and parallel processing with FPGAs are moving these examples from the drawing board to the runway, as well as provide flexibility for future alterations, which is particularly important when dealing with data streaming in from multiple sources, not to mention evolving industry standards.

The problem is, FPGA development can be very complicated. Not every team has the knowledge base or time to dig into system design and the tedious verification tasks that accompany that work.

But it doesn’t have to hold those teams back.

With strict size, weight, and power (SWAP) requirements, advanced embedded technology is crucial to aerospace systems.

A company that develops imaging solutions for aerospace applications came to DornerWorks for FPGA design and IP implementation, for an advanced driver-assistance system (ADAS) that optimized data ingestion from a dozen simultaneously streaming cameras.

At the start of the project, DornerWorks organized its hardware, software, and FPGA teams to provide the client with system level expertise, and fill knowledge gaps.

The multi-camera sensor imaging system for data collection DornerWorks developed for the customer involves five separate cameras networked via HDMI, CameraLink, CoaxExpress, and other imaging interfaces.

Custom logic built for this system provides:

  • Video Synchronization and Triggering
  • MetaData insertion
  • Color correction and contrast enhancement
  • Image sharpening and transformation

Through a simple, iterative approach, DornerWorks engineers developed a working demo designed to withstand a range of scenarios as well as pair up with existing development platforms. The client then used that prototype to secure new revenue streams with a wider portfolio.

The FPGA-based system provides the company with flexibility in a field subject to rapidly evolving standards and requirements. And, despite having little to no prior FPGA experience, the company found a way to develop a design that would bolster its reputation in the market, without having to redesign a single aircraft.

Matthew Russell
by Matthew Russell