Driver’s behavior

By Albert Zakhia, November 2017

Harsh Acceleration/Deceleration debunked

Today, there is a high demand for driver’s behavior. As a matter of fact, the global market is being geared toward driver’s behavior.

Still, the market is lost with the misconception that “harsh acceleration/deceleration as provided by current GPS devices” are the solution.

Although harsh driving is a part of driver’s behavior, by itself it means nothing.

In this article we will show part one of our study and explain what is required for correct interpretation of driver’s behavior and show how harsh driving is part of the whole.

We will also show why individual companies cannot implement driver’s behavior by themselves due for its complexity and high cost. For this, private companies will need to turn to government agencies to help research, fund and offer such a service.

Driver’s Behavior

Today, most if not all tracking companies are basing their driver’s behavior from what is well known as

  • Harsh Acceleration
  • Harsh Deceleration
  • Harsh Cornering

All 3 events are based on Acceleration

  • Harsh Acceleration = positive rate of change of velocity in time
  • Harsh Deceleration = negative rate of change of velocity in time
  • Harsh Cornering = high rate of change of heading in time

We will generalize all 3 harsh accelerations as Harsh Behavior instead of Driver’s Behavior. This is so because harsh behavior is one aspect of a large set of other parameters required for calculating driver’s behavior. By itself it cannot be considered as such.

Acceleration as used in current tracking devices is based on the following formula:

The above formula shows that acceleration is the difference of velocity (speed) within a given time interval. By Harsh acceleration, we understand a high difference in velocity within the same given time interval. This is the formula used by Tracking Device Manufacturers to calculate harsh accelerations/deceleration.

This FORMULA FAILS for the context of our requirements!

It fails because it does not take into consideration a major multiplier: mass! This means a mosquito that hits you while accelerating 1 meter per second2 will have the same impact on your body as a car hitting you with the same acceleration. Try it!

Taking mass into consideration, getting hit by a mosquito accelerating 1 meter per second will not do any damage, whereas being hit by a 1 TON vehicle is at the very minimum, going to put you in hospital. With a 30 TON truck, you are dead! Same for vehicles, a Porsche accelerating from 0 to 90 in 3.4 seconds is somehow ok and fun, whereas a truck with a full load of explosive should not reach 80 and if it does, should do so in not less than 60 seconds. Hence vehicle type, usage, weight, cargo and many other parameters play an essential role in determining correct harsh acceleration and parameters.

Hence, harsh acceleration as is has no true value in driver’s behavior. For the same truck, if the device is configured for full load (higher mass) then it will provide false acceleration positives when it is empty (lower mass).

For the same truck, if the device is configured for empty load (lower mass) then it will not provide us with any event when it is full (higher mass) because it is impossible to have the same optimal acceleration for both loads (empty and full).

In order to circumvent that, the device should be configured with the correct parameter both on load and unload; that is partially impossible and the client has already wasted his money.

Yet, harsh acceleration will prove useful once integrated with other data. This is the subject of our presentation.

Viewer discretion advised

In the above video, (and many other) we can see many potentials for driver’s behavior.

  • FOG: in such circumstance, driver should have decreased speed in first sign of fog and low visibility; this was not the case.
  • Most vehicles had their lights and warning lights off (before and at time of the accident).

Similar accidents will repeat again as a result of rain (low visibility, hydro gliding), sandstorms (null visibility) as well as many other conditions.

What solutions?

First comes “prevention“. Drivers should be trained to the latest preventive standards. In addition, training should be periodic. A Driver should understand that he is the captain of his vessel, and hence, before embarking on any journey, he should be well aware of:

  1. Weather condition: different weather conditions require different speeds. If a highway has a max speed of 100 km/hour, under fog the speed should drop to 50 or 60 km per hour, depending on visibility. If there is rain and possibility of hydro gliding, then maybe the speed should drop even more.
  2. Road condition: from source to destination. Different road conditions require different speeds. (gravel, sand, marked, unmarked, multilane, ascending, descending, curves, visibility due to objects, animals crossing, traffic…)
  3. Vehicle’s condition and its capability of reacting properly to both conditions as per points 1 and 2. (fluids, gas tank over half-full, plenty of windshield washer fluid. Belts, hoses, and brake systems checked for excessive wear. Exhaust system checked for CO leaks into passenger compartment. Tire treads and correct pressure checked for adequate traction. Windshield wipers blades replaced if ineffective.
  4. Cargo type and condition, making sure the type of cargo can handle point 1, 2 and 3. Hazmat (Hazardous Material) cargo requires more preparation and lesser speeds for safe transportation. Acceptable truck speed for water tanks are essentially dangerous for trucks carrying explosives.
  5. His own physical/emotional condition: clean from alcohol (sober), having full rest from previous journey, the ability and readiness to cope with all 4 points for the totality of the journey.

Followed by “monitoring”. Drivers should be monitored to make sure they are using the vehicles in a safe way. Examples:

  1. Alcohol sensor to detect alcohol in driver’s breath is an option. The engine should not turn on in case of alcohol traces. We need to study existing crash statistics to figure out if it is worth it to invest in such a solution.
  2. Distance sensor that will monitor the driver maintaining a safe distance behind a facing vehicle.
  3. Fatigue/drowsiness facial sensors which senses the driver’s facial expression for fatigue
  4. Excessive hours driving. Drivers should rest for a small period every given number of hours.
  5. Mobile application that will lock the driver’s mobile at speeds higher than 20 or 30 km/h hence disallowing the driver to use his mobile while driving.

Implementing the solution

This part provides a small description of the solution. Interested parties willing to take this further and invest in the solution will receive the solution after signing an NDA.

A brief of the solution:

  1. Gather data country wide
  2. Centralize it
  3. Force drivers to respect it

Phase 1: Gather as much data as possible from the ground. Client uploads their trip information as soon as they receive them to a central server. Accidents together with the road conditions at the location of the accident and the vehicle’s state at the time are recorded. Given all this information, the system can start learning and providing baselines for every type of truck and ground condition. Clients will get recurrent and automated reports for their drivers.

Phase 2: Live collected data can be pushed to every driver’s mobile based on his location together with the road condition max-speed baseline hint. ‘Warning, fog ahead, slow down..’ or ‘Slippery surface ahead, slow down..’ message-like will be received. The driver has to slow down or he will keep receiving such warning messages and an overspeed will be logged.

Phase 3: Drivers will be categorized based on their driving habits. Any company needing a driver can have a look at any driver’s log and, based on his driving score and history, can decide whether to hire him or not. This will force drivers to be better and accumulate points. (Lost points are accumulated with time)

This is a magnificent project to invest in. After proving its efficacity, the more drivers registered the more precise the baseline will be. Forcing drivers to remain within the baseline will help decrease accidents. Accidents when they later happen will contribute in fine tune the baseline..

And so it goes..