Holy Grail of Safety Systems: Zero Fatalities

BRIMLEY, MI – Someday, our descendants may look back in head-shaking amazement at the days when cars collided and killed people.

Today, skeptics struggle with the prospects of auto safety technology reaching the nirvana of zero-free fatalities.

Ibro Muharemovic is no such doubter. A Continental engineer, he embraces the supplier’s vision of a world without traffic deaths.

Existing advanced-safety technology, such as collision-avoidance and automatic-braking systems, already make modern cars less accident-prone. Future systems will eliminate accidents altogether, he says.

Muharemovic is the lead engineer for the German auto supplier’s work on an autonomous car. It uses radar, cameras and other sensors to drive itself.

Born in Bosnia, he moved with his family to Germany in 1993 and to the U.S. six years later. He went to high school (L’Anse Creuse) and college (Oakland University) in Southeast Michigan, where he now lives (Shelby Township). He joined Continental as an intern in 2005.

During a preview of pre-production products at Continental’s winter-test track here outside Sault Ste. Marie in Michigan’s Upper Peninsula, he speaks with WardsAuto about the development and dynamics of advanced-safety systems and how they protect drivers, sometimes from themselves.    

WardsAuto: In developing products, does Continental try to anticipate what automakers want or do you try to create a market to sell in it?

Muharemovic: Our Chassis and Safety (Div.) is very data-driven. We look at all the accident databases. There were 33,000 traffic fatalities in the U.S. last year. We pull the numbers and see that 15,000 of them involve people leaving the roadway.

What can we do about it? Here are our sensing systems, here are our suspension systems. Let’s combine things for smoother integrations. That’s often how we come up with our projects.

Sometimes people say “That’s great, it’s really going to help us.” Other times they say “That’s nice, but…” If it’s a hit, we’ll go forward with it. If it’s not, we’ll come up with something new or change it up or integrate it differently.

WardsAuto: So you study accident data?

Muharemovic:  We have teams looking at it and trying to find correlations across a broad spectrum in fatalities. You try to group them into categories. You have this big group of drunk drivers. What can you do against that? One thing is to have the car manage it better.

WardsAuto: What about driver-distraction accidents involving texting and cell-phone use? That must be a relatively new equation you are dealing with.

Muharemovic: It is. The thing is, you can’t really control what the driver does. For us, the critical point is recognizing what he’s doing. Is he distracted? What is his attention level? What is his intention? What is his cognitive state? Is he in a foul mood or a good mood? Is he aggressive or not?

Based on all that information, we build a driver-attention model. As we showed today on the track, we can warn early or late, automatically brake early or late, depending on a system determination of the driver’s attention level.

Gauging Driver Mood

WardsAuto: How do you determine driver mood?

Muharemovic: There are different approaches right now. You can (use cameras to) look at the eyes and eye movement. You can look at behavior with the steering wheel and brake pedal on a daily commute. You know how a drowsy person drives, drives and then oops, there’s a correction. Drives, drives, oops, there’s another correction. You kind of determine this guy may be headed in a certain direction.

We do a lot of studies with individual drivers because everyone has different driving styles. You need to get a lot of statistics. We give people cars to drive so we can get the data. Based on that, you have your factors. Universities also collect and analyze massive amounts of data for us.

WardsAuto: The systems look at deviations in someone’s daily driving, and if something is different, the systems will react to that?

Muharemovic: Not deviations from the daily driving, just the driving of the moment. We don’t store anything. We just look at something like: OK, every 10 minutes or so he’s doing this or that, and it gets worse.

WardsAuto: Has the integration of systems always been going on or is it a new way of thinking?

Muharemovic:  In the past, you had, say, your brake and engine suppliers. The engine guys were interested in accelerating, the brake guys in stopping.

Now that we have all these components, there’s not much more to add. We can make existing stuff better, but we have to use what is there. So let’s combine them into real systems.

We take what we feel is the next step to reduce accidents and combine that into an enhanced system. It varies from project to project. It’s free hand.

WardsAuto: What’s that mean?

Muharemovic: It means we determine a use case where, say, a certain percentage of rollover accidents occur. We say, “We have the brakes, we have the suspension. We can combine our active-safety information with active suspension and braking to prevent rollovers.” It’s up to us to determine how to use systems in our portfolio to address driving issues.

WardsAuto: What does the data say is the biggest cause of accidents these days?

Muharemovic: Drunk drivers are No.1. About half of fatal accidents are from people leaving the roadway.

WardsAuto: And most people who are likely to leave the road are drunk?

Muharemovic: We know more than half the fatalities are because people left the road.

WardsAuto: So you don’t necessarily need to be drunk to do that?

Muharemovic: Not necessarily. We all are trying to reduce fatalities. We have this vision of accident-free driving with no fatalities. It’s really a core.

WardsAuto: You mention it as a vision, but will it become a reality?

Muharemovic: Of course. We’re putting a lot on the line to make that happen.

WardsAuto: Is that possible, practically speaking?

Muharemovic: We think so. That’s why we’re here. If we come close, that’s good. But we can always get better.

WardsAuto: But won’t there be a drunk driver of the future capable of killing himself or somebody else, no matter what advanced-safety features are on the car?

Muharemovic: We hope we can develop automated driving systems that handle everything. We have to try. We owe it to ourselves. With that many fatalities a year, we have to do something.

WardsAuto: What are drunk drivers doing specifically to be the main cause of fatal traffic accidents?

Muharemovic: Not being in control of the vehicle. Not being able to judge speeds, lanes, oncoming traffic. They don’t know where to go, really.

WardsAuto: And they end up going where?

Muharemovic: Often into another car or into a person. But if, for example, we can make pedestrian protection work (with a system that recognizes pedestrians and automatically brakes for them if the driver doesn’t), that’s another set of fatalities we reduce. That’s what we aim to do, and what we must do.

[email protected]