Building better DAQ tools.
About this time, Vrancic transferred internally from NI's DAQgroup to its LabVIEW Real Time group--LabVIEW is NI's graphical operating system for control and measurement devices. While there, Vrancic helped develop a timed-loop scheduler for the LabVIEW RT structure, while also fixing some Ethernet driver issues. This work led Vrancic to propose a solution for replacing memory-reflection cards with NI's off-the-shelf Ethernet cards that would become a deterministic Ethernet system.
Once the timed-loop scheduler was released, Vrancic starting working on the deterministic Ethernet infrastructure, which was released as a part of LabVIEW release 8.0. The combined work on the timed-loop scheduler and deterministic communications infrastructure resulted in a new LabVIEW Real Time component that is now known as an ns-engine. Ns-engine is a component that takes care of all software timing within a system. It can be synchronized to a GPS (global positioning system), an onboard clock that has a much more stable frequency. As a result, a user can fire a synchronized timed loop in a distributed system, and when the GPS support is added, these systems can be located on opposite sides of the world and still function.
While still working on the original timed loop project, Vrancic started toying with the idea of using timed loops as a way to target code to different CPUs. On Windows, the OS (operating system) infrastructure was there, but. more was needed on the Real Time system. So, over time he started making small incremental steps until NI purchased Phar Lap OS code. This accelerated his work, and within a few weeks, Vrancic was working on prototypes. He now is working on a Real Time SMP (symmetric multi-processing) solution.
Vrancic describes his job responsibilities as consisting of three parts--the investigation of new technologies and methods, the definition and delivery of final products, and determining ways to mitigate hardware-software interactions. Part of his job is also attempting to hide as many of these issues from the customer as he can during development.
During the nine years that he's been a part of the NI development process, Vrancic has seen his workload increase, but finds that he doesn't have any additional work, because he's become much more efficient at solving the problems he's given. He finds that he's now mostly self-driven in the projects he works on because he's become more aware of the complexities and issues than anyone else--that's what experience is all about in a research environment.
Vrancic's biggest challenge is balancing the R in R&D--which is what he'd like to do--with the D in R&D--which is needed to create the products that drive NI's revenue stream.
When asked what he is most proud of, he lists his work on contributing to his company's LabVIEW Real Time systems. "LabVIEW will become an awesome real-time platform," he says. "You can always program some things in fast OS like C, but you won't have the flexibility that you have with LabVIEW."
Vrancic's vision of his future is based doing much of the same work that he does now, while continuing to learn more, experiment more, and follow the ever accelerating technology curve. Of course, he still finds the time to help build robots with his 8-year-old son. But since it's mostly based on real time systems, that too could be considered an integral part of his job--looking for future applications.
Principal Engineer National Instruments Austin, Texas
* BS Electrical Engineering, 1991. Univ. of Zagreb. Croatia
* PhD Physics. 1997. Louisiana State Univ.
* Engineering Excellence Awards from National Instruments in 1999, 2002. and 2006
* Multiple U.S. Patents
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Ethernet; data acquisition|
|Publication:||R & D|
|Date:||Sep 1, 2006|
|Previous Article:||World-class technologies for R&D and beyond: the 44th Annual R&D 100 Awards honor the world's best innovators and their innovations.|
|Next Article:||Composite optics advance stellar imaging: composite materials are making it possible to add capabilities to astronomical observatories without...|