Innovation is alive and well in today’s engineering communities as evidenced by the trove of design ideas sent to Electronic Design for publication by readers, vendors, staff, and other contributors. It is incredible to see the true levels of imagination and engineering represented in each submission. And it seems that no matter how many new ideas come along, there is never a shortage of new problems in need of innovative solutions.
This is equally true in test and measurement as it is in design. Test engineers are experiencing a plethora of unprecedented challenges in need of innovation across all industries. In the communications industry it’s about keeping up with the latest wireless standards from 3G to Long-Term Evolution (LTE) and now LTE-Advanced. In the automotive industry we are immersed in the expansion of alternative energy-powered vehicles. Aerospace and defense is undergoing a wide assortment of transformations due to the changing financial, political, and military environments.
Two primary drivers of innovation in test and measurement are the needs to solve critical business and technology-oriented challenges. Critical business challenges include the demand for increased productivity, faster time-to-market, increased reuse and return on investment (ROI), lower total cost of ownership, and increased throughput. Increasingly complex technical challenges in test and measurement include the growing software-defined nature of devices under test (DUTs) and their associated software-defined test systems, multiple models of computation, heterogeneous processing and types of hardware targets providing increasingly faster I/O, rapidly emerging commercial technologies, and integrated system-level timing and synchronization for starters.
National Instruments recently detailed an innovative business and technology-aligned approach, called graphical system design, to help foster innovation in test, measurement, and control at our annual NIWeek user conference in Austin, Texas. For systems that need measurement and control, graphical system design is an approach to visualize and implement systems with an open system platform of software and hardware. It incorporates technology in a way that abstracts system complexity, making new technology more easily accessible to engineers in their solutions.
Abstraction Of Commercial Technologies
Figure 1 shows how a graphical system design platform can abstract the complexity of FPGA technology to the pin. In this example, a clear graphical representation of system functionality replaces thousands of lines of equivalent VHDL code. The same platform also abstracts programming complexity when taking advantage of multicore processors or DSPs. Any other commercial standard technology, including communication technologies and protocols, are abstracted in the same manner.
With this abstraction of complexity, engineers can focus on using technology to serve the end goal of the system, whether it’s a control system, a test system, or an embedded system. Without this approach, engineers seeking a competitive advantage through better performance and lower costs of commercial technology either have to focus on programming components or interface with specialists who can do so. Both options increase time-to-market.
Graphical system design accelerates development by sufficiently abstracting complexity yet providing access to the pin, making it possible for engineers to easily use technology for the purpose of the system. With the graphical system design approach, engineers can gain the performance and cost benefits of commercial technology without increasing time-to-market.
Integration Of Multiple Software Methods
Graphical system design enables engineers to explore multiple ways of solving a problem and lock in on the best options more quickly than traditional methods. As engineers visualize system functionality, different system components may need different methods, or models of computation, to best describe that functionality. For example, parallel programming is best represented as a graphic, but equations are represented well using text. The structure of the system may be state-based, sequential, parallel based on dataflow, or a mixed model (Fig. 2).
Graphical system design software can incorporate multiple models of computation, or methods, within a single exploration space to give engineers access to the best methods for the functionality they need. In doing so, graphical system design abstracts complexity at the system level, where system components of different models can be placed together in a single software platform space, and then integrated together visually, functionally, and in an architecturally sound way.
The graphical system design approach includes both software and hardware as part of the exploration and implementation platform. Often engineers have excellent high-level software tools. But when they get to the real implementation of prototypes or end systems, the tool chain starts slowing down development. Models either break down at boundary conditions, they don’t behave as expected in the real world, or there are difficulties in implementing the software in the hardware.
With graphical system design, the single, open-platform approach integrates software with customizable off-the-shelf hardware platforms from beginning to end. This approach takes a comprehensive view of the system for the end purpose of functioning as a controller for wind turbines, an automated test system for cell phones, or a robot that can perform surgery on humans.
Every system that needs measurement and control requires hardware implementation. Graphical system design includes the same platform elements in multiple hardware form factors, making it possible for engineers with customizable off-the-shelf options to explore solutions. The common hardware platform elements—processing (processors, DSP, FPGA), communications, and modular I/O—are abstracted at a system level in the same way as abstracting models and other software constructs. Once engineers invest in the learning curve of a system platform, they can integrate and iterate quickly at each stage of the product development cycle.
Graphical System Design In Practice
At the extreme edge of this problem is next-generation development in aerospace and defense. Recently the Defense Advanced Research Projects Agency (DARPA) announced funding research for the META program, which seeks to improve and accelerate development of defense systems that incorporate software and electromechanical systems through model-based design methods. The graphical system design approach scales much better with these requirements than traditional methods because it abstracts the complexity of system components, both software and hardware, and focuses on the integration of those components.
The productivity benefits of the graphical system design platform span every industry in which engineers are creating systems that need measurement and control. At BioRep Technologies, engineers use graphical system design to control complex automated medical instrumentation (Fig. 3). The platform approach to system design provided them with a single learning curve for software and hardware, reducing their development time from a year to three months.
Graphical system design scales to solve automated measurement systems, where the software platform abstracts the elements of the I/O, analysis, communications, and visualization needed to upgrade or integrate instrumentation functionality. Texas Instruments, for example, used graphical system design for its automated test platform and reduced test time by 70%, while increasing test coverage by more than 100% for production testing of power-management ICs.
Ian Wong of National Instruments also used graphical system design to showcase the world’s first public prototype implementation of an 8x8 multiple-input multiple-output (MIMO) LTE-Advanced physical layer (PHY) on a commercial off-the-shelf platform, achieving a data rate of close to 1 Gbit/s. The hardware platform comprised a high-performance quad-core Intel i7 multicore processor, NI FlexRIO FPGA processing modules with Xilinx Virtex-5 FPGAs, NI FlexRIO adapter modules for baseband digital-to-analog and analog-to-digital conversion, and PXI RF modulators and converters for over-the-air transmission (Fig. 4).
This high-performance platform, coupled with LabVIEW system design software and the LabVIEW FPGA Module, helped a small team of communications and signal processing algorithm engineers at NI prototype a complex LTE-Advanced communication system in six man-months. This project was part of advanced research National Instruments is doing in communications system design. For more about this solution, see “Getting Ready For 8x8 MIMO.”
Reconfigurable Hardware Keeps Up With Moore’s Law
Graphical system design is a reconfigurable system approach for developing systems that need measurement and control. Using this approach, engineers can invest development through a single platform from which multiple solutions can be derived to serve a broad range of application needs. As the platform incorporates new technology, systems derived from it can use that technology and gain the performance and cost benefits of Moore’s Law for systems.
Scientists at the Instituto de Astrofísica de Canarias discovered this benefit while developing a system to position actuators for the European Extremely Large Telescope Array. The initial performance tradeoff they expected to make by going to a customizable off-the-shelf platform as opposed to custom design never materialized. In fact, they exceeded their performance requirements while significantly reducing development time.
Platform Ecosystem Drives Innovation
Engineers can leverage the work of other engineers using the graphical system design approach. One of the great benefits of platform-based development is the ecosystem of intellectual property (IP) and applications that are regularly shared within the community. In the world of computing and mobile devices, the innovation that’s driven around platforms is obvious.
PC, Linux, iPhone, and Android are all supported by tremendous ecosystems of developers, IP, and applications. In fact, the adoption of virtual instrumentation in the test industry, which was supported by the ecosystem of PCs in software and I/O peripherals, is one example of how the graphical system design approach revolutionized the ability for engineers to gain significantly higher performance at lower costs faster for a specific application space.
Many engineering tools and approaches focus on optimizing specific areas of the design-to-implementation process, even while painting the picture of an integrated flow. Often these are platforms where either software or hardware takes the primary focus and the integration of the overall system is de-prioritized. Engineers can easily select tool chains that continue to require more expertise and time to get to the real solution.
With graphical system design, engineers gain an open, reconfigurable platform that accesses multiple technologies and tools at a high level in a way that tightly integrates software and hardware, shortening the most time-consuming portion of the design process. An open platform also generates an ecosystem of IP so an individual or small team of engineers can leverage years of development without investing additional resources. While many engineers have already had considerable success with this method, graphical system design remains an accessible opportunity for many looking to gain disruptive competitive advantage.
Innovation In A New Era
With all new eras of tools, approaches, and expertise over time, it is important to keep the core business and technical requirements for the changes at the forefront to ensure the maximum return of your investment. Graphical system design was created to help you overcome the increasingly complex task of developing highly integrated test, measurement, and control systems using the latest industry technologies and open standards without a large team of dedicated experts in each area. The result is most often a much faster time-to-market, easier to maintain modular systems, lower total cost of ownership, and the maximum available performance of your systems based on the use of the latest technologies available on the market.