The Human Factors Field Research Vehicle: FHWA Takes Its Show On The Road

The year is 1998. It's a rainy winter evening, and you are struggling to read the handwritten, cryptic directions to the holiday dinner at your boss's home.

The rain-slicked lane markers are nearly impossible to see, and your less-than-perfect eyes are alternating between monitoring the frustrated commuter behind you, who is exhibiting classic manifestations of road rage, and straining to read road signs that resemble the sixth line of the eye test in your doctor's office. The 80-year-old driver in front of you isn't helping matters any, as he too struggles with the elements and periodically hits the brakes with seemingly little rhyme or reason. And added to this already troublesome mix are an array of road signs dictating high-occupancy-vehicle restrictions, warning of road construction ahead, and a variable message sign with the words "Accident Ahead, Expect Delays."

The good news is that you finally reach your boss's home. The bad news is that dinner is over and the other guests have left.

Flash forward to the year 2000. Once again, it is a rainy winter evening. Only this time you are desperate to reach an unfamiliar hospital to which your child was taken after a crash on the way home from an out-of-town basketball game.

The Human Factors Field Research Vehicle is a four-door 1995 Pontiac Bonneville packed with about 360 kilograms (800 pounds) of computers, sensors, display panels, video cameras and recoreders, microphones, and other technologies to measure older driver performance in real-world driving situations.

Fortunately, an automated navigation system in your dashboard, using Global Positioning System (GPS) data from satellites, is delivering turn-by-turn instructions to guide you to the hospital. Notice of an accident ahead prompts you to ask the onboard navigation system for an alternate route, and within seconds, it has established a new route for your journey. As the onboard computer instructs you to turn right at the next intersection, a collision-avoidance monitor alerts you that the confused teenager in front of you is braking hard for no discernible reason. Fortunately, a vision enhancement system is making it much easier to read the dimly-lit street signs and lane markers.

The good news is that you reach the hospital quickly and discover that your child is not seriously injured. The bad news is that it's not the year 2000 yet, and the technology that safely delivered you to the hospital isn't quite ready to become "standard equipment" on automobiles.

But, if the engineers at the Federal Highway Administration's (FHWA) Turner-Fairbank Highway Research Center (TFHRC) have their way, these technologies soon will make driving an easier, more efficient, and - most importantly - a safer means of transportation.

The Research Vehicle
In a laboratory at TFHRC, a handful of human factors engineers/research psychologists are conducting a variety of experiments that they hope will one day fundamentally alter the way people drive. By studying individual drivers and the way they react to a number of external and internal stimuli, FHWA plans to develop and evaluate new technology that facilitates the driving experience.

Part of this process is the Human Factors Field Research Vehicle (HFFRV), a four-door 1995 Pontiac Bonneville packed with computers, sensors, LCD (liquid crystal display) panels, video cameras and recorders, microphones, and assorted other technologies. Known euphemistically both as the "Veda car," after the company that outfitted it for FHWA, and "RealSim" by the engineers who work with it, HFFRV is just one of four laboratories in FHWA's Human Factors Program. The others are a sign simulator; a part-task driving simulator; and a fully interactive, high-fidelity highway driving simulator.

As its name denotes, RealSim is a test vehicle that permits FHWA engineers to take laboratory experiments to the field for real-world testing. Comparisons can then be made between data collected in the simulators and actual driver responses in real-world driving conditions.

"Simulators are always going to be limited to some extent by their inability to incorporate the incredible variety of real-world conditions," says Spencer James, a research psychologist working on the project. "Even the very best simulators can't account for all of the things we find in the real world - all the distractions and events that catch the attention of the driver."

Enter RealSim, which permits James and his colleagues to replicate many of their in-house experiments on the road. A focal point of that research centers on RealSim's reconfigurable dashboard, which enables researchers to physically arrange and rearrange the location and relationship of dashboard displays to evaluate driver response times, fatigue load, preferences, and much more. For example, by reconfiguring the LCD panels, researchers can determine the relationship between display timing, information priority, display location, and information presentation and performance. In other words, the positions of the displays can be tailored to the driver's needs and requirements.

"One of the things we're looking at is providing essential information to the driver in a manner that does not compromise safety," says James. "If you've got an 80-year-old driver, is he going to become confused or freeze at the information you are providing him? Does he want that information presented on a pop-up display or in the dash? Or with a 16-year-old driver, is he going to get caught up in all the gadgetry and drive into a tree? There's a lot to consider, not just in the technology, but in the requirements of drivers of all ages and temperaments."

Like jet fighter designers, who struggle to avoid overloading a pilot by paring down the amount of information he or she receives, today's transportation planners must also consider the volume of data that threatens to inundate many drivers facing rapidly changing driving conditions and congested, often deteriorating roads.

Making the Car Smarter, the Driver Safer
"The modern driving experience is vastly different from what it was even 10 years ago," said Kathryn Wochinger, a research psychologist with Science Applications International Corp., which is working with FHWA on the RealSim project. "You've got enormous congestion, frustrated drivers, often-times poor signage, and a rapidly growing population of elderly drivers who have special needs."

The trick, according to Wochinger, is to provide drivers with the right amount of information, at the right time, and in the proper presentation mode.

"We want to reduce congestion and improve efficiency, but not at the risk of compromising safety," she said. "The key is to do all three things at the same time."

But that is easier said than done. For example, an elderly driver is likely to have slower reaction speeds to warning sensors. But he or she is also apt to be more easily alarmed or confused by a sudden infusion of information. Younger drivers, on the other hand, are prone to overconfidence and might find themselves obsessed with dashboard instrumentation to the detriment of their own safety.

"You don't want the guy crashing into the car in front of him while he's marveling at the technology. It kind of defeats the purpose," James said.

"Reaction times vary considerably from one age group to another," Wochinger said. "Ultimately, what we are trying to do is create systems that improve all drivers' abilities. If we can do that, we can reduce congestion on the roads and improve safety."

To examine these and other issues, RealSim has been outfitted with a host of commercial off-the-shelf technology designed to study driver habits. Onboard equipment includes five personal 486 computers, which power a variety of components, including the experimenter's station system (ESS), data-acquisition system (DAS), driver response panel (DRP), video data-acquisition system (VAS), in-vehicle display and control systems (IDCS), navigation/map system (N/MS), and a lane-tracking system.

To one degree or another, all of this equipment allows researchers to collect a wide variety of data, including lane deviation, speed, the position of the automobile via GPS, acceleration/deceleration rates, verbal and manual driver responses, and driver physiology - for example, tracking the driver's eyes to determine what he or she is observing and doing while certain road and other real-world conditions are being encountered.

Determining a driver's ability to recognize and comprehend verbal cues and various visual icons on dashboard displays.
Assessing the best position for in-vehicle display data, including head-up or in-dashboard displays.
Measuring a driver's acceptance and use of in-vehicle safety warning systems.
Determining the degree to which in-vehicle information systems result in information overload to the driver.
Evaluating different instrument layouts based on driver preferences and requirements.

The Nuts and Bolts of RealSim
With more than 360 kilograms of equipment onboard, the Bonneville's rear springs were replaced with heavy-duty shocks and springs designed for recreational vehicles. Similarly, to keep all that technology humming, the vehicle's alternator and regulator were replaced with a heavy-duty 190-amp truck alternator and an external 14VDC regulator.

Up front, the entire instrument panel was removed to accommodate four LCD panels for the reconfigurable display. Analog sensors were added to the steering wheel, accelerator, and brake pedal to measure driver inputs. An interface also was added to the engine's control computer to measure vehicle performance. GPS and DGPS (Differential GPS) antennas were added to run the map display system and record the time of the videotape recordings. Six video cameras were installed throughout the automobile to record both the subject and vehicle situational data.

The experimental systems themselves contain three major components: an experimental setup and control system, an in-vehicle display system, and a data-acquisition system. Each component employs open system architectures and is modular in nature, so that changes can be made to individual components without reconfiguring the entire system.

Experimental Setup and Control System
From the back seat of the vehicle, the experimenter uses the ESS work station (which includes a flat-panel VGA monitor, keyboard and mouse, video monitor, and video switch) to run various tests as RealSim travels Northern Virginia's roads. This in-vehicle work station is not to be confused with the off-vehicle station, which is run out of the TFHRC Human Factors Laboratory. Using an "events list" that details each action to be taken during the test, various scenarios are programmed and then, for comparison purposes, reloaded (using a custom programming language) into the in-vehicle ESS for real-world tests. Programmable events are based on simple algorithms - for example, "If speed exceeds 55 miles per hour [90 kilometers per hour] at location X, initiate audio."

In-Vehicle Display System
This system employs the hardware and software to display information on the five LCD panels in the reconfigurable display. It comprises three distinct subsystems: displays, instrument panel graphics, and a navigation/map system. The LCD panels can be used to display a variety of information in graphic or symbolic form, including standard data - speed, fuel, temperature - and data related to intelligent transportation systems - collision warning, in-vehicle signing, turn-by-turn routing.

Auditory information is generated via digital voice or prerecorded messages through the ESS work station in the back seat, and the data is saved to an audio file that can be accessed and played back for postexperiment studies. Video and audio displays can be triggered in a variety of ways, including latitude/longitude positions (via GPS), odometer readings, and speed. Speakers are located behind the driver's seat.

The displays employ a PC-based graphics-generation software and special graphics-display buffer boards. Real-time data from the vehicle is used to trigger symbolic displays, which are developed via Avionics Visual Instrument Development Station (AVIDS) software. For example, the speedometer symbology is driven by the speed data contained on the vehicle's data bus.

The trunk of the Human Factors Field Research Vehicle is also packed with testing and recording equipment.

Testing and recording equipment in trunk.

A 264-millimeter LCD panel is home to the navigation/map data. The panel employs a touch screen that allows control of such functions as map, traffic, or route planning (turn-by-turn) information. Using data supplied by the DGPS antenna and receiver, the nav/map system provides real-time navigation information to the driver. The event list, which is developed before the test run, allows the system to display different routing and geographic information on the moving map. As a result, the experimenter can evaluate the driver's response to various events, such as changing a route due to a simulated accident. Each new route can be triggered and displayed via time, distance, location, or experimenter inputs.

Data-Acquisition System
This system has three subsystems: the driver response panel, the computer data-acquisition system, and the video data-acquisition system. These systems record human and vehicle performance data, including eye/head movement, lane tracking, and audio. All of the recorded data is time-stamped for post-processing correlation. The heart of DAS is the five shared memory cards that create a unified hardware communications scheme. The cards provide memory for the application program to read from and write to, and all five cards are instantly updated when one of the values is changed. Recorded data is archived to a removable hard disk for in-house analysis.

The driver response panel is located between the right arm rest and the transmission shift, and it contains one large push-button momentary switch, three small push-button momentary switches, three analog linear potentiometers, and two five-position rotary switches. Data is also collected from the driver via eight momentary push buttons mounted on the left and right of the steering wheel. All of the aforementioned buttons and switches can measure a driver's state of awareness and workload while operating the vehicle.

The computer data-acquisition system uses 16 analog input channels to register driver and vehicle response times. Standard sensors register a number of variables, including accelerator position, brake-force pressure, steering wheel position, fuel level, water temperature, oil pressure, three driver-response potentiometers, and a three-channel accelerometer.

The video data-acquisition system employs six cameras: one each mounted on the trunk for right and left lane tracking, one in the dash and another above the dash for tracking driver eye and head movement, and two behind the driver for capturing window views. All video data is captured on six Hi-8 videotape recorders located in the trunk. The video signals are routed through a vertical interval time code generator, which synchronizes time and position information onto each video frame for postexperimental analysis.

Making Sense of It All
"It's a lot of equipment," said James, eyeing the technology-packed Bonneville as it was prepared for a road test. "But so much goes on during the average drive that it takes a lot of gear to capture it all. And it's important that we analyze as much data as possible to determine which technologies will make our roads safer and less congested."

"It's important to remember that we're not trying to create in-vehicle technology that will replace the driver's actions," Wochinger said. "We're developing technology to complement his abilities - to make it easier to read those signs; to find the quickest, most efficient route to a destination; to avoid collisions; to remain in his own lane. These are all important efforts because they all contribute to more efficient transportation, less congestion, and safer roads."

The Bonneville's engine starts; the LCD panels light up; and RealSim emerges from its home at TFHRC for another road trip. It's a trip the researchers hope will one day make all our lives easier and safer.

Doug Rekenthaler Jr. is a freelance writer and editor. His experiences as a writer and editor include cub reporter covering Capitol Hill and Pentagon news beats; managing editor responsible for 12 newsletters covering a wide array of communications technologies; founder of the multimedia industry's first daily fax news service; and corporate communications manager for America Online Inc., the largest commercial online service in the world.