Paul Green


Paul Green in Office

Meet Paul Green

“The most important things that affected whether the ship was effective were human factors issues.  Anyone could design a foundation for a boiler or figure out what size pump was needed, but nobody could figure out the human issues…this brought me here to Michigan.”~ Paul Green, PhD


Dr. Paul A. Green is a research professor in UMTRI’s Driver Interface Group and an adjunct professor in the University of Michigan (U-M) Department of Industrial and Operations Engineering (IOE).  He is also a past president of the Human Factors and Ergonomics Society.  Dr. Green teaches automotive human factors and human-computer interaction classes.  He is the leader of U-M’s Human Factors Engineering Short Course, the flagship continuing education course in the profession, now in its 55th year.

Dr. Green leads a research team that focuses on driver distraction, driver workload, and workload managers, navigation system design, and motor-vehicle controls and displays. That research makes extensive use of instrumented cars and driving simulators.

Dr. Green’s research has been published in close to 300 journal articles, proceedings papers, and technical reports.  He was the lead author of several landmark publications: the first set of U.S. DOT telematics guidelines and SAE recommended practices concerning navigation system design (SAE J2364, the 15-second rule) and design compliance calculations (SAE J2365).  He is the lead author of SAE J2944, Operational Definitions of Driving Performance Measures and Statistics.

Before joining UMTRI, Dr. Green was an engineering staff member at a co-op job with the Philadelphia Naval Shipyard and a safety and health engineer for Scovill. At U-M, he has held appointments in the Departments of Psychology and Mechanical Engineering, the School of Art (Industrial Design), and the School of Information.  He has a B.S. degree in mechanical engineering from Drexel University and three degrees from U-M: an M.S.E. in Industrial Operations Engineering (IOE), an M.A. in psychology, and a joint Ph.D. in IOE and psychology.  He has taught every semester for 28 years as adjunct.

 What spurred your interest in this research?

As an undergrad student, I entered a fuel economy contest at Drexel and became interested in role development and auto instrument panels display and driver interface.  The field that I work in, which is human factors (works in Driver Interface group) is important because for a lot of the technologies that we are developing the engineers, the mechanical engineers, the electrical engineers, they will do their jobs, and they will build systems and they will work from a functional perspective but they are not necessarily useful to people.  And obviously that focus on the human element is really a cornerstone of the work that happens here at UMTRI.  If you think of all the groups that work here, I’d say there are three pillars to the work at UMTRI.  There is safety, vehicles, and people.  UMTRI to its credit has had this long focus on the human side of driving.

What is your focus now?

On driver interfaces and continues to be related on driver workload and predicting what drivers do.  The latest buzzword is driver distraction.

I am currently working on warning systems in cars, basically using the question, how should you warn the driver of various things that are related to crash avoidance and their response to lights and sounds and its factors (specific designs and placements).

My second project is funded by National Science Foundation to look at intersection warning and things you can’t see.  We built a simulation of people going through many, many intersections and we’re now looking at augmented reality to warn drivers.  Augmented reality is…imagine your entire windshield could be a warning display.  One scenario is where you are approaching an intersection and there is a car off to the side, it’s actually hidden by a building, but it’s going to collide with you and you can’t see it… first we have this little arrow that says there’s a car over there, then when the car comes in view we have a little bar that tracks the car, but you wouldn’t see it otherwise. The way it would be implemented technically is you would have a little image on the windshield and would track the vehicle. So first thing is how you warn a driver about something they cannot see and that’s the most valuable warning to a driver because if you’re warning about something they already know about, it’s a not a very informative warning.  The warning has to be reliable enough.

The ATLAS project asks how do we predict the time that it takes drivers to do tasks that are distracting? The notion is the longer it takes, the more distracting it is. For example, it’s fine to adjust the speed for the fan to the climate control because you can do that pretty quickly.  It’s not a good idea to go and search your song database for a song title especially when you have to do something like typing in the name of the song; it just takes too long. That’s a distraction.  So the notion is that you want to allow short tasks and to not allow long ones.  So how do you determine the duration? The federal government has in the NHTSA visual manual guidelines… something called an occlusion test. This is one of two ways that they specify whether or not a task is too long. It’s an expensive procedure in looking for ways to predict time.  So the procedure is as follows:

Occlusion goggles

You wear a pair of goggles with liquid crystal shutters in them…the lens is either milky or clear. They switch between seeing the road and not seeing the road and seeing the display  and not seeing the display—meaning it’s trying to simulate performing a task while you drive.  The lenses simulate real-life behavior by alternating between being clear and milky for a 1.5 second period which is roughly the duration of a glance.  For example, if it takes more than eight glances to enter a destination into a GPS, it’s too long a timeframe, and that what this research is all about.

So we’ve been working on methods to predict performance in these tasks by really looking hard at what the steps a person does and trying to reduce the number of steps.  This project is trying to reduce the amount of steps, or pushes of buttons or number of screens; trying to make it so the task is constructed in a way that they can do it in less time.    We want to encourage automobile engineers to give plans six months ahead of production so that the human factor elements can be incorporated and changes can be made well ahead of the manufacture of the hardware and software.

The important design decisions will be made six months to a year upstream based on these calculations; the goal of this research. What I think is important about this is that a lot of the human factors work is just let’s go run a bunch of subjects and see what happens. But, in this case we’ve got the data from those kind of evaluations, what’s important to have is calculations and predictions well ahead of time.  The hidden advantage is then this approach becomes like all other engineering.  If you want to put together a circuit to do something, you don’t say let’s go throw a bunch of resistors and capacitors together and see what happens; you have Thevinen circuit equations and Kirchoff’s Voltage Law, and you predict and you know what the voltage of the current is going to be.  So our philosophy is understanding enough about people so that we can build these laws, rules, and equations, and when somebody says what do you think about this design, we can say here’s the performance that we can predict from a person doing this task.  That’s the way all other engineering works.  The advantage is that when you present these results to engineers they’re much more likely to accept them.  If it’s more like engineering, and we show things like the rest of engineering, then they are much more likely to accept it and think about the human element as being just another block in the diagram and then they are much more likely to do the right thing.  While if it’s something they don’t understand, and is unpredictable, an unknown, they will ignore it.  That’s the real problem. If you don’t know about it, then ignore it. While if it’s predictable, then okay.   Humans are just another engineering compound that just happens to be alive.

So it’s as much about the specific technical issues…what we deliver makes driving safer, but it has larger implications for how do we think about doing human factors and building the systems in which people are components. To quantify what the person does. That’s really the key.

Human factors isn’t just about studying how people are driving, but it’s the perspective we take to lots of things.  It informs us in valuable ways.

What is the future of your research?

Predict performance with warning systems that don’t exist. This will be technically feasible in the future.  “You want a driver to get into a car they have never driven before, experience a (standard) warning for the first time,  and know what it is and what to do about it immediately…even if they’ve never been exposed to the technology.”

What impact will your research have on society and what are you working toward in your research?

The key to the ATLAS work is that we want to make it so that engineers can compute the extent to which something is going to be distracting before it’s ever built. We need to do that by allowing people to compute or estimate the extent to which something is going to be distracting. And in this case, as specified in the NHSTA protocol.

The vehicles being produced in the future will be safer because the extent to which potentially distracting systems are implemented will be less.

Tessa Elwart - Student

Meet Tessa Elwart

Undergraduate Student Research Assistant

Tessa is a transfer student from Washtenaw Community College and started in research through the U-M Summer UROP Research program.  Her interest in cars stems from her dad working at GM. She is entering her senior year at U-M and will graduate with a B.S.E. in Mechanical Engineering December 2015. One of her leading interests is people.  She loves helping and interacting with people which ultimately led to her interest in research.  It seemed to be a perfect match for Tessa to work on a project at the University of Michigan Transportation Research Institute (UMTRI) in the Driver Interface Group.

The current project she is working on has to do with estimating task times while using a navigation-radio.  These task times determine if that task is excessively distracting.  This information can be used to change the design of navigation-radios to become less distracting and reduce distraction-affected crashes.

After graduating, Tessa plans on getting a job to gain more experience and will ultimately return to graduate school with a better idea of focus for her area of study. Tessa is a member of the flag line in the Michigan Marching Band and coachs volleyball for Huron Valley Volleyball Club.  In her free time, she loves playing volleyball (see video below), walking in the woods, and spending time with her friends and family.


The ATLAS Center is a collaboration between the University of Michigan (U-M) Transportation Research Institute (UMTRI) and Texas A&M Transportation Institute