Electronic Travel Diary
The Psion unit employs an object-oriented programming language (OVAL) which is similar to Visual Basic. The programming language is complex, but very powerful. The object-oriented program allows the provision of onscreen commands that allow users to turn from survey page to survey page at the push of a button. Hence, rather than scrolling through pages on the screen, larger pages can be broken into nested pages (where a button click takes the user from one page to another). The nested program capability also reduces the number of screens with which a user must interface. Screens that are not applicable, based upon previous data input, are never displayed.
The research team did not examine alternative handheld computers that employ touch screen technologies. Touch-screen units were outside the scope of the project, in part because a successful touch-screen unit was already in use on a FHWA project and alternatives needed to be evaluated. Unfortunately, Sony no longer sells the Magic Link, employed in the FHWA/Battelle travel diary project in Lexington, KY. Thus, the only touch-screen unit currently programmed for use as an electronic travel diary is unavailable for future studies (the replacement Sony technology requires new programming). Under a separate research project, Georgia Tech researchers are currently reviewing the specifications of several touch screen devices that came on the market in 1998. If any of these units appear promising for providing superior performance over the Psion unit, they will be procured and tested at Georgia Tech. GDOT and ARC will employ the Psion or an alternative unit in instrumented data collection activities associated with the year 2000 Atlanta household travel survey.
Global Positioning System Options
After evaluating all measures of performance for each of the GPS units, the GeoResearch Workhorse appears to give the best overall performance. One important factor for the travel diary study is the unit's ability to consistently report data in one-second intervals. Another factor is that that unit provides good coverage of the street segment without missing large portions of data. It is also important that the reported data match correctly to the corresponding street segments since ultimately the data from individual vehicles will be matched to a street network. The Workhorse meets all of these requirements and in most cases outperformed the other units. For any GPS unit, the temporal accuracy issue must be addressed in combining data streams.
The Garmin 35LP performed well, especially given the uncorrected data. Because it is available at a much lower cost than the Workhorse, it may be considered for studies in which the results are not as dependent on the unit's ability to consistently provide high positional or temporal accuracy. For example, the use of the lower cost receivers would be more suited to an application such as a travel time study, where the street segments traversed are known and runs with missing data may be discarded. In addition, the Garmin unit will support real-time signal correction if a correction signal is available in the study region (either broadcast by the researchers or provided by a third party). If a correction signal is available for a study region, the Garmin units may be able to achieve accuracy comparable to the GeoResearch Workhorse.
Real-time differential correction may now a feasible alternative for use in electronic travel diary studies in Atlanta. Many GPS units can now be equipped to receive a radio signal broadcast directly from a base station by the research team or by a private company that operates a local base station. The correction signal compensates in real time for the selective availability introduced by the military. Because a correction signal arrives at the field GPS unit, there is no need to retain the copious satellite data required for laboratory post-processing. In essence, only the x, y, z, time, and a few precision measures need to be retained by the GPS. The reduced need for handling data can significantly reduce system memory requirements, power draw associated with recording data, and post-processing labor. The research team is currently field-testing two real-time differentially corrected GPS units in Atlanta as a component of a separate research project. These units are compatible with the Psion handheld diary. Researchers will use a Garmin 35LP GPS with real-time differential correction in Atlanta's year 2000 travel survey update, provided the real-time correction signal is viable throughout the region.
Current Onboard Diagnostics Unit
The Snap-On MT2500 engine analyzer outputs a data stream of engine parameters. The interface is proprietary and logging of data is accomplished through an existing DOS-based program. This program is manually operated to start and stop the logging of the data stream. The program was also limited in the length of time it allowed logging. The research team worked with Snap-On software developers to determine the potential for modifying the Snap-On program to remove the time limit on logging the data stream and to allow control from a LabVIEW program. A quote was generated for the modifications and held until funding could be identified. In the intervening period, other customers began asking for Windows-based interfaces. Snap-On developed a PC translator program to communicate with the Snap-On Scanner through one serial port, translating the engine parameters to text and transmitted them out through a second serial port. This new program interfaces the Snap-On unit to other equipment in automobile test cells.
Snap-On is currently modifying the new program into what they are calling a dual-port RAM interface. The program will initialize the Snap-On unit, allow selection of engine parameters to report and then log these parameters into a dual-ported memory area that can be read by other programs. The interface reports the parameter names and units as well as the values. It also reports a special code to indicate engine-off. From the description given, it seems that this program could meet the qualifications to interface the current control program with the Snap-On scanner. This interface program is scheduled to be available in March 1999. The research team will continue to investigate the usefulness of the Snap-On scanner as new software becomes available.
Meanwhile, the OTC unit is the only viable onboard engine scanner available at this time. The reliability of the OTC to PC serial link was good during development testing. A few instances of OTC unit lockups occurred when initialization keystrokes were sent too quickly. In these instances, a power-off reset was required. One basic problem with the unit is that a manual keystroke is required to begin the data collection process. If the unit were to lose power during a testing period, the unit cannot restart without human interaction.
The OTC unit only provides second-by-second data when used in non-OBDII mode. Tests with new vehicles using the generic OBDII module resulted in 3-second data resolution. This resolution is insufficient for use in developing speed/acceleration profiles for use in engine load analyses. The lack of a viable standard OBDII mode for this unit increases setup labor during equipment installation. Parameters must be selected manually and, in some cases, adjustments to the LabVIEW code may be required. The research team believes that the same communication method is used for standard OBDII and proprietary data streams on newer vehicles, so there is currently no way around this limitation for newer vehicles. The team will continue to work with manufacturers of scanning equipment and urge them to improve their software such that 1Hz resolution will be achieved.
The AED VCS PCMCIA card provides acceptable data for OBDII-equipped vehicles. Six to seven engine parameters can be sent to the computer every second. The PCMICA unit also consumes 2/3 of the power that is consumed by the OTC. However, since completion of the analyses, the research team was informed that AED has gone out of business. The research team is currently searching for other PCMCIA-based scanning tools.
Current Durable Portable Computer
Datalux Databrick II
Although the prototype system incorporates the Databrick II, power draw issues and temperature concerns have prompted further research of rugged laptops possessing low power suspend modes and high temperature tolerances. AED's Roadrunner was the best option found, with temperature tolerances ranging up to 70oC and with a deep suspend mode having minimal power draw. Price discussions were conducted with AED during the December 1998. However, AED went out of business as of January 1, 1999 and thus the Roadrunner is not available for future use.
The next best solution is CyComm's PCMobile, which has a 586 processor, a monochrome monitor with a detachable keyboard, temperature tolerances up to 60oC (140oF), and a deep suspend /low power draw mode. Discussions on functionality and price have been completed with CyComm and the research team is currently waiting for the allocation of funds to purchase and evaluate this laptop within the instrumentation package. As a result of this change in the computer component, new casing will also be acquired to allow for the necessary clearance of a laptop monitor and for cooling and venting options.
Copyright 1998-1999 GeorgiaInstitute of Technology