The Internet of Things technology trend is influencing the strategies of companies throughout the world. The combination of internet connectivity with embedded computers and sensors is leading forward-looking businesses across different sectors, from industrial process suppliers to consumer devices, to put IoT into their devices and business models.
These “converged devices” combine mechanical, electrical, telematics and communications elements to provide functionality. The need to mix these various technology elements is having a massive effect on the design and development cycle of products and devices, not least in the area of testing and validation. Perhaps the foremost example of the converged device is the driverless car.
Dr James Truchard, chief executive of test and measurement equipment company National Instruments, says: “It has to be proved that driverless cars will be safe. We have the platform that can make that happen. Our ability to test converged systems is important. We can work with the embedded systems at the same time as we work with things like radar.”
The company is supporting the development of IoT-enabled devices with several of its latest hardware releases and updates to the Labview software used to program them for test and measurement tasks, says Truchard. Crucially, the company’s Hardware in the Loop (HiL) simulators enable engineers to more easily and quickly develop the “brains” – the computers embedded in things such as cars and aircraft to run the sensors and collect and display data. The use of a HiL simulator enables more testing and development of embedded systems to be carried out virtually in the laboratory, instead of using physical testing.
Detecting bugs in radar
A HiL simulator, in combination with National Instruments’ second-generation vector signal transceiver (VST), is being used in the testing of automotive radar for Audi’s latest cars. Audi and its partner Konrad Technologies have built a test system that can provide the type of bandwidth a radar needs with the lowest latency possible, so it can simulate other vehicles and buildings. It is being used to detect critical bugs in the radar system.
As with all car manufacturers, Audi’s latest driverless system has to meet ASIL D requirements. This applies to both the hardware transceiver and the software.
According to Michael Konrad, chief executive of Konrad Technologies, the most difficult part is testing the software, because it requires testing over a massive number of different scenarios. “To validate an ASIL D radar system, the car has to drive more than 10 million km, 250 times around the world. It would take more than 10 years,” he says. “So to efficiently test it we have to recreate the environment using a target simulation.”
The HiL simulator allows engineers to emulate 10 years of sensor environment in just a few weeks. The simulator uses National Instruments’ second-generation VST because “the larger bandwith gives us better precision,” says Konrad. The software allows for the simulation of Doppler shift in the signal, to recreate radar bouncing off objects. “The software allows us to recreate any type of pattern,” he adds, tricking the sensor to think it is seeing objects in real time such as lane changes and objects entering its path.
Troubleshooting
Niels Koch, component owner of radar systems at Audi, says: “The combination of the industry’s widest bandwidth and low-latency software-designed instrument allowed us to discover our radar sensors as never before, and even allowed us to identify problems early in the design phase that were previously impossible to catch. With the VST and field programmable gate array programmable by Labview, we were able to rapidly emulate a wide range of scenarios, thus influencing safety and reliability aspects in autonomous driving.”
Konrad adds: “The dream of fully autonomous driving requires technology with sensors fusing a global navigation satellite system, radar, lidar and cameras. We expect to tune our system to combine more advanced sensor technology in future.”
The possession of equipment to deliver this type of converged device is just one aspect of IoT-focused product development. Another critical element is the expertise to use such equipment. There is a growing need for engineers to cross the boundaries of traditional disciplines, with projects requiring knowledge of electronics, wireless signals and software programming. This requirement has already begun to influence education.
Some education providers are tailoring courses to enable students to make complex, multidisciplinary products faster and better. At the University of Southampton, undergraduates focus on the rapid prototyping of wireless communications systems, again using NI equipment – in this case a universal software radio peripheral (USRP). This transceiver turns a standard PC into a wireless prototyping system, allowing live radio frequency signal streaming.
Tracking aircraft
For example, students set up a USRP to receive the automatic dependent surveillance – broadcast of real commercial aircraft, obtaining the position, heading and tail number of the aircraft. They then used NI Labview with a Google Maps plug-in to track the aircraft, and validated their positions with the Flightradar website.
Rob Maunder, an associate professor at the University of Southampton, says that most communications laboratories are not properly equipped to train engineers to deal with the challenges presented by wireless communications. Such training requires technology such as HiL simulators and USRPs.
“We can offer something distinctive – not every university offers HiL teaching,” he says. “The benefit for students is a boost in confidence. It enhances their employability. Hyper-connectivity is driving developments in consumer and industrial technology. We need to bring the IoT into educational laboratories.”
Andy Bell, director of academic programmes for NI, says that a deeper understanding of test and measurement tools that provide data and insight during design and development can be enabled through hands-on projects with students. “Networked objects are producing unimaginable amounts of data, creating new challenges. We must take the next step in education, innovating and designing in new ways,” he says.
“We want to accelerate discovery, moving down this path from insight to implementation.”
SPOTLIGHT – CROSSRAIL IN THE LOOP
It’s not just the automotive and aerospace sectors that are adopting Hardware in the Loop (HiL) simulation to speed up development and testing. London’s Crossrail trains have also been developed using the technology.
Inspired by the aerospace industry’s use of “Iron Birds” to optimise vital systems at the earliest opportunity in the development cycle, Bombardier and Frazer-Nash consultancy developed the Train Zero facility, to try out model-based design techniques.
Train Zero can optimise everything from requirements capture and validation, through de-sign, and on to validation and verification testing at a sub-system and system level. HiL and integration testing is carried out far earlier in the development cycle than was possible be-fore in the design of rolling stock. It provides early confidence in development and promotes commonality across different systems. The efficiency of testing has been increased and validation of trains has been streamlined.
The system uses National Instruments’ Veri Stand and PXI modules for testing of the train systems and validation of the models, so changes made to the models can be revalidated in the environment they will later be running in. Colin Freeman, senior consultant from Frazer-Nash says: “NI’s development platform provides the versatility of data acquisition hardware that Bombardier required, as well as a software environment capable of executing the variety of models it was working with in real time.”