In 2012, there were 1.7 million rear-end crashes.

That’s what a special report by the National Transportation Safety Board (NTSB) revealed. The number represents almost half of two-vehicle crashes, and what’s most shocking is that these crashes resulted in 1,705 fatalities and more than a half million injuries. Why? Reasons include driver inattentiveness, fatigue, and reduced visibility. In short, drivers could not identify and respond to conditions quickly enough. As engineers, it is our job to seek solutions to problems like this.


Can more sensor technology integrated in vehicles help us avoid accidents and save lives?

For students to fully understand what is required to answer this question, they need to understand the current market. Already, sensor technologies have increased vehicle safety. Automobiles include rear backup cameras, lane change detectors, and occupancy detection sensors. They serve as the eyes and ears of the vehicle (but with better information), and they capture conditions at higher speeds than humans. Entrepreneurs are already developing new vehicular sensors. However, there is a fundamental challenge. When an automotive maker integrates any new technology, it needs to work reliably for a large customer base and gain widespread acceptance. The technology must create value for the customer and within the larger context of society and the economy. Successful solutions will increase safety, have a clear customer value proposition, consider the ethics of man-machine interaction and safety, and meet applicable regulatory requirements. I introduce students to many promising technologies — one of them being radar. Radar systems rely on radio frequencies in the electromagnetic spectrum in the millimeter wave band. For instance, Delphi Automotive offers a bumper-mounted electronic scanning radar (ESR) that uses a solid-state technology with no moving parts and operates in the 76 GHz spectrum. Digital signal processing (DSP) algorithms are applied to process the radar echo signals for the detection of static and moving targets to calculate their range, range rate, and azimuth angles. In mid-range mode, Delphi’s ESR radar has a range of 60 meters at a field of view of + 45°. In the long-range mode, its detection range can extend up to 174 meters but at a narrower field of view of + 10°.

I want to spark curiosity in my students and demonstrate how radar technology has the potential to save lives. Since radars are expensive sensors for everyday classroom use, I utilize a low-cost HC-SR04 ultrasonic sensor to demonstrate the same concepts. The sensor offers a simple digital interface to an Arduino-type microcontroller. It has a sensing range of 2 cm to 400 cm at a resolution of 0.33 cm, with an angular field of view of about + 15°. It operates with a 5V supply, drawing 2 mA of quiescent current, and a typical operating current of 15 mA. This sensor uses a short 10 microsecond (us) pulse applied to the trigger input to start the ranging. Then the transmitter unit in the ultrasonic sensor emits a burst of eight cycles of 40 KHz pulses, and raises its echo pin. The echo pin remains high until an echo signal is received back from the target, at which point the signal goes low.  The pulse width of the echo signal is proportional to the target distance.

Here is how students derive the formula for the distance calculation:

Screen Shot 2017-04-06 at 9.54.40 AM







D: distance between the sensor and the target
T: pulse width in microseconds (us). This corresponds to the round trip travel of the ultrasonic signal from the sensor to the target and echo back to the sensor.
V: speed of sound in air (343 m/s = 34300 cm/s in dry air at 20°C or 68°F)

Applying curiosity and making connections:

To understand the limitations of this sensor in high speed automotive applications, I ask students to analyze its properties, including its range and update rate. They determine if the sensor is appropriate for crash avoidance in high speed operations.

Exercise 1:

What is the maximum update rate for the HC-SR04 ultrasonic sensor at its maximum detection range of 400 cm? Compare this update rate to that of Delphi’s ESR radar, which is capable of updating every 50 ms.

Exercise 2:

Assume that an ultrasonic sensor mounted on a vehicle for obstacle detection application has a maximum detection range of 30 m. Let us assume that, on average, a vehicle traveling at 45 mph has a braking distance of 30 m. Braking distance is the distance a vehicle travels from the point when its brakes are fully applied to when it comes to a complete stop. Taking into account the limited detection range of the sensor and the time it takes for the ultrasonic sensor to read the range measurement (i.e. the time of flight measurement of the ultrasonic waves), determine if this sensor will be suitable for obstacle detection and crash avoidance application for the specified vehicle speed and braking distance.

Exercise 3:

A human operated vehicle has a stopping distance that is made up of two components: perception-reaction distance and braking distance. The perception-reaction distance is the distance the vehicle travels from the instant the driver perceives a condition on the road that needs a reaction; decides what maneuver is appropriate (such as stopping the vehicle); and takes the action (such as applying the brake). Advanced Driver Assistance Systems (ADAS) provide automatic functions to detect hazardous conditions and alert the user or automatically take corrective action to avoid accidents. Such systems help significantly reduce the perception-reaction time compared to that of human beings, who take on average from one to two and a half seconds depending on how alert the driver is. Distracted drivers, drivers under the influence, and the elderly have relatively longer reaction time. The total stopping distance for a vehicle can be expressed as follows:

Screen Shot 2017-04-06 at 10.01.06 AM





Where V is the speed of the vehicle before braking is applied, m is the coefficient of friction on the road surface, and g is the gravitational acceleration. Assume m =0.7, and g = 9.8 m/s2. Given the above explanation and equation for stopping distance of a vehicle in response to potential accidents, work through the following problems in teams: Assume that a vehicle is driving at 60 mph. If the driver’s reaction time is 1.5 seconds, how far will the vehicle travel before the driver reacts? If the reaction in the aforementioned problem requires applying a full brake, calculate the braking and stopping distance of the vehicle. Now assume that the vehicle is equipped with ADAS that has an automatic braking feature to prevent accidents. If we assume the reaction time for the ADAS is 100 ms, calculate the braking and stopping distance of the vehicle.

Exercise 4:

Suppose a vehicle is using an ultrasonic sensor with a maximum detection range of 30 meters for its ADAS. Calculate the maximum vehicle speed that you will be able to use this ultrasonic sensor for crash avoidance, assuming the reaction time for the ADAS is 100 ms.

Students demonstrate their curiosity by going beyond the automotive safety applications that are the main focus of the classroom activities. Once exposed to the components that comprise stopping distance, students extend the concept to other applications. One team of students discussed ideas on the use of ultrasonic sensors for industrial safety and automatic door control applications. By thoughtfully decomposing things that matter into components, they are demonstrating their curiosity. Students make connections between the ultrasonic-based robot and the opportunity to create a real-world solution for collision avoidance. They also integrate technical information about the stopping distance of vehicles and the sensor characteristics, such as its measurement range and update rate. This helps them appreciate how engineers need to consider the constraints of the sensors. By the end of the class, students not only have the technical understanding of why vehicles equipped with sensor technology will help avoid accidents and save lives, but they are also equipped with marketplace knowledge to navigate how such innovation can and should be implemented. In order to create safer and better vehicles, we need engineers with an entrepreneurial mindset to make it happen.

Practice Makes Perfect

Screen Shot 2017-04-06 at 10.14.30 AMIn the experimental activities of my class, students use a small robotic car equipped with an ultrasonic sensor for collision detection and avoidance. Students first experiment with the use of the ultrasonic sensor for distance measurements. They develop a program for implementing obstacle avoidance behavior of the robot using its ultrasonic sensor. To simulate a vehicle, the robot travels in a straight line until it encounters an obstacle. When it detects an obstacle within a given range (such as 30 cm), the robot executes an obstacle avoidance maneuver to steer itself away from the obstacle and then it continues to move in a straight line. Once students demonstrate the basic skills in using the ultrasonic sensor, an additional challenge is provided. I ask students to implement a line-follower algorithm (using light sensors covered in a different course module) and the ultrasonic sensor for collision avoidance and basic cruise control capability. In the basic cruise control mode, the robot is expected to adjust its speed according to its distance to other vehicles in front of it.