Novi, MI. Data acquisition, data management, HiL test, and environmental test were all subjects under investigation at this week's Automotive Testing Expo. Presenters covered these and related topics in the event's Open Technology Forum, while the presenters' companies highlighted related instruments, systems, and software on the exhibit floor.
Test data acquisition and analysis received the most attention from presenters. Jordan Van Baren, domestic sales manager of Vibration Research, described how to make the most of your field test data in the lab. He described the traditional field tests of the '80s and earlier, when a test vehicle would be tethered to a van full of rack-mounted test equipment that ran a course parallel to the test track, acquiring real-world data that could be replayed on a shaker table in the lab. Today, acquisition hardware has shrunk considerably.
Van Baren advised making sure to collect enough data: you can always down-sample later—you cannot up-sample. He also described various sources of error: signal-conditioning circuit anomalies, interfering electromagnetic fields, resonances, power-inverter noise (if you are running your equipment from a test vehicle's DC supply), and cable whip (the triboelectric effect).
Once you've acquired field data, he said, you need to decide how to play it back in the lab. You may want to play the data back exactly as it was acquired, although that can take significant amounts of time, and you may want to edit the data—to remove dead spaces, for example. Alternatively, you might want to create a random PSD (power spectral density) or fatigue damage spectrum.
Raimund Trummer, director of product marketing for Dewetron in Austria, addressed requirements for automotive data acquisition. He described a DAQ “automotive toolbox” that can manage DAQ chores for combustion analysis, power analysis (for electric and hybrid vehicles), road profiling, adaptive cruise control, pedestrian protection, lane-keeping assistance, cross-wind stabilization, and the many other functions of a modern automobile. His company, he said, offers all-in-one DAQ systems, mainframes, and front-ends (which work with an external computer for data processing). Trummer described in particular the company's SYNC-CLOCK technology, which can operate autonomously or disciplined to, for instance, an external GPS signal. He also described the company's Polygon product, which, he said, customers have referred to as the Swiss Army knife of vehicle test. He presented a forward collision warning (FCW) example. He advised that engineers performing such test acquire human-sensible data—use a microphone to capture an audible warning, for example—don't rely on a CAN bus signal.
Jim Schwarz, product marketing engineer at National Instruments, commented on in-vehicle datalogging. Cars traditionally had few or no ECUs, he said, but now they have tens or hundreds. Consequently, cars are increasingly software-defined. A modern vehicle, he said, may have a hundred million lines of code. Engineers still need to acquire and log traditional sound, temperature, pressure, and vibration data, but they must also log data on CAN and LIN buses. Tools that can help, he said, include the NI CompactDAQ platform and NI DIAdem analysis software.
On the exhibit floor, NI highlighted its Drivven subsidiary, which presented a Smart Car outfitted with CompactDAQ and CompactRIO platforms. Drivven is using the platform, along with other NI products such as NI DIAdem, in an effort to improve the efficiencies of internal-combustion engines.
Although Van Baren of Vibration Research advised making sure to collect enough data with the option of down-sampling later, such an approach could ultimately be counter-productive. That's the conclusion I drew from a Open Technology Forum presentation from G. Sammer, global product manager for AVL in Austria, titled “Handling the Data Explosion from Gigabytes to Petabytes.” He noted that exponential increases in complexity are leading to corresponding increases in data volume, and data is expensive, he said. IT staffs and data storage itself won't keep pace with the increasing data volumes, he said. He recommended a seamless integration of data formats from office simulations to road tests, with a common data backbone. He recommended ways of rejecting data that's not usable, consistent, or plausable (perhaps because it was acquired by a faulty sensor) and that's not productive (temperature sampled at 100 Hz, for example). He cited AVL's Santorin Host as a system that can help manage data, and he concluded that responsibility for the relevance of acquired data lies with DAQ engineers, not IT departments.
Russell Wallace, who is responsible for technical services at imc DataWorks, offered a presentation titled “Effective HiL-based Testing Utilizing Synchronized Measurement, Simulation, and Control.” His company's focus is testing, he said, but its product line does offer support for modeling (and can run Simulink models) and can handle real-time calculations and signal generation. “You can simulate what you don't have,” he said. He presented an example of an airliner wing flap test. The manufacturer build one prototype wing and simulated the other to verify that the flaps would operate in a synchronized manner.
Addressing environmental test was John Komar in a presentation titled “Creating the Perfect Storm in the Name of Innovation.” Komar is the director of engineering and operations at the Automotive Centre of Excellence (ACE) at the University of Ontario Institute of Technology. ACE, he said, has been operating a little over a year and makes available $100M worth of environmental test tools government, industrial, and academic researchers. ACE, he said, can order up four seasons in one day at the flip of a switch. ACE, he said, serves applications ranging from automotive test to TV production, where directors may need to film a wintry scene in August.
On the exhibit floor, Agilent highlighted its 34972A DAQ system, its PXI-based TS-8900 automotive functional test system, the enhanced B1505A power device analyzer/curve tracer (which operates to 10 kV and 1500 A), the Agilent 33500B Series waveform generators (with Trueform signal-generation technology), the 100-MHz U1610A and 200-MHz U1620A handheld oscilloscopes that include a color VGA display, PXI-form-factor chassis and test modules, and a selection of bench and handheld instruments, including Agilent Infiniivision 3000 X-Series oscilloscopes, modular power systems, handheld instruments, and USB modular instruments.
As I post this article, Johnnie Hancock, Agilent product manager, is slated to offer an Open Technology Forum presentation titled “Characterizing the Differential CANbus Using Oscilloscope Eye-Diagram Mask Testing.” You'll find his comments on automotive bus test in our October feature “Instruments, Software Drive Serial Bus Success.”
See related article “Mild Hybridization Could Boost Automotive Efficiency.”