Two of the NAE grand challenges are centred around health; engineer better medicines and advance health informatics (Engineeringchallenges.org, 2018). Using data collected from 2016, around 181,384 casualties are reported to the police each year, with more than 660,000 road accidents occurring nationwide. (Assets.publishing.service.gov.uk, 2018) With the number of vehicles on the road increasing, ensuring drivers are safe is a pressing issue. A study done using EMS Patient Care Reports and police crash reports for the entire state of Alabama (Gonzalez et al., 2009) show a clear correlation between increased time passed between the accident and the hospital, and increased mortality rate. Using the data from rural areas and excluding any patients dead on scene, the mean response timer mortalities is over two minutes longer than for those patients who survive.
In my project I have decided to look at implementing an automatic response of ambulances to car crashes. This would include detecting the crash with sensors on the car to show any rapid deceleration or force on the bumper, asking the driver to indicate if they were okay, and if not an ambulance would be sent immediately to the GPS coordinates of the car. My main objectives are to design the digital circuit and code that would implement this program, as well as research the need for this technology, particularly in rural areas.
Accident statistics
In 2016, car occupants (whom this technology is aimed at helping) accounted for 46% of the total road accident fatalities in that year. If we can help reduce this number by reducing the prehospital time, this would save both lives and money.
From a study undertaken in the US, around 75% of casualties occurred in rural areas. However when looking at the proportion of mortalities, this rises to 86%. The driving is often more dangerous in rural areas as there are narrow roads and the areas tend to be less well-lit, explaining the larger proportion of casualties. The hospitals also tend to be more spread out in less populated areas, so the travel time for the ambulance to respond is often longer. This makes it even more vital that they are called immediately. Finally, I also believe that in urban areas, accidents would be identified immediately, whereas in rural areas where less people drive along the roads, someone could have crashed and no one would discover them for hours at a time. This is one of the most important uses of this technology, as the ambulance could be called immediately regardless of anyone witnessing the crash and the car could be located with its GPS coordinates. Reducing the time between the crash and help being provided by professionals with sufficient equipment is the most important thing in reducing mortalities, as shown in the introduction.
Sensors to detect the crash:
There are many different types of sensors used in cars to detect crashes and deploy airbags or to measure whether the car has flipped. In the past, the most common sensor used was a simple pressure sensor that detected the impact of the crash, however now the technology is moving towards detection of the deceleration caused by the crash. Multi-axis sensors can detect horizontal deceleration caused by impact on the front of the vehicle as well as vertical acceleration (when the car rolls) caused by impact on the side of the vehicle. (Kaur, 2018) This is economically beneficial as well, as only one sensor is required.
An excellent example of a multi-axis sensor is the central acceleration sensor designed by Bosch. It can be integrated with the airbag deployment, and will detect crashes from the front or side of the car. To make the application of this technology more feasible to distribute widely, it is easier to simply use whatever sensor the car currently uses in its airbag-deployment detection as this will already be fitted inside the engine.
Circuit design:
Accident occurs → Accident detected → User check → Contact ambulance
There should already be a multi-axis sensor in the car that would be integrated with the airbag system. This is an example used at the front and side of the car to detect impact with another object. When this occurs, the sensing plate is forced forwards onto the contacts, and creates an electrical connection between the two terminals, completing the electrical circuit.
When this circuit is completed, this output needs to be tested to check whether is is above a threshold voltage to determine if the crash is severe enough. This would be done using an op amp as a comparator. This works by comparing the voltage produced with a reference voltage and checking to see which one has a larger magnitude.
Using a non-inverting comparator circuit, while the input voltage is smaller than the reference voltage the output stays at zero. However as soon as the input voltage exceeds the reference voltage then the op amp goes into saturation. Here the output voltage is equal to the difference between the input terminals multiplied by the open loop gain of the op amp.
This covers the first two steps of the flow chart above, and as this technology should already be available in most cars, this will be relatively cheap to implement.
User input:
Within this technology, I would like to ask the driver to press a button to confirm they are okay even if an accident is detected and the airbags have been deployed. This would reduce any unnecessary calls to the emergency services which would be detrimental to other patients needing help. If the button is pressed, the process would be reset, however if it is not pressed it would continue onto contacting the ambulance. This would involve putting buttons or another kind of sensor into the car, or connecting the circuit to something already available in the vehicle.
Control section:
Assuming no input from the user, the output from the detector would then be passed to a micro-controller. I would choose to use a micro-controller over a microprocessor because it is a more compact component and also has RAM embedded in the system, allowing the data to be read quickly, whereas the microprocessor would require additional external circuits. The device I will be using is a 16F84a PIC, as it is suitable for beginners and available for simulations on Multisim. The name tells us it is a flash device and can work up to a 84Hz oscillation frequency.
The most common power supply used for microcontrollers is the 7805 3-terminal voltage regulator I.C. (Smith, 2007) with a supply voltage between 7V and 30V. There are some components that are essential to the microcontroller circuit, such as a capacitor placed between 5V and 0V to allow electrical noise to bypass the microcontroller and the crystal and capacitors connected to pins 15 and 16 that produce the clock pulses required for the circuit. Inputs and outputs are then connected to the microcontroller via the I/O pins.
The first input that the micro controller will receive is the output voltage from the op amp comparator. This is a digital signal so should be connected to a digital pin on the microcontroller.
Contacting the ambulance:
This part of the circuit will require a GSM module (global system for mobile communications) to be connected to the microcontroller. This allows the microcontroller to contact the emergency services with the use of a GSM modem containing a SIM card.