1. Equipment
A self-driving car would be equipped with many sensors to provide information about road conditions in order to control its speed, brake, steering wheel as well as safe distance with other objects or cars on the roads. Those equipment could be
- GPS with precision to lane
- Photo sensor with image processing or pattern recognition capability
- Radar or laser beam to estimate safe distance between the car and other objects
- Weather forecast center or the information control dashboard should be connected to a local weather forecast station via wireless (mobile router of a cellular phone) in order to know the current weather such as raining or snowing.
Figure 1. A self-driving car with its sensors
2. Laser beam
The car would be equipped with 8 laser beams in order to estimate the distance between itself and other objects in front of, behind, and around it.
There are many laser tools to estimate distance of an object and a laser tool in the market.
The laser beam would help to estimate, or control speed of a car, or its position (moving a little bit to the right or left) for a safe distance on the roads with other objects around it.
There is a physic formula to calculate distance based on the bounced back signal strength of a wave signal, i.e. used by laser beam in this case.
If the car in front of it is driving at slower speed than the upper speed limit suggested by the GPS, the car could try to change lane to pass that car.
The front/rear of the car could be equipped with 3 laser beams each to detect objects in the front/back of it.
In this note, the laser beams and photo sensors were used to determine safe distance. For changing lanes and other tasks, more laser beams and sensors would be needed.
3. Photo sensor or capture
The front/rear of the car could be equipped with 3 laser beams each to detect objects in the front/back of it.
In this note, the laser beams and photo sensors were used to determine safe distance. For changing lanes and other tasks, more laser beams and sensors would be needed.
3. Photo sensor or capture
a. Laser beam failed
There is a case that the laser beams would fail, i.e. snowing or raining days.
In the case that the distance reported by the laser beam changed drastically, the photo sensors would be activated to confirm the weather around the car and estimate safe distance along with the laser beam.
The photo sensor is special that it could estimate if an object similar with a car’s shape is so close to itself in order to slow down the vehicle.
b. Rain fall and snow fall
The rain fall and snow fall could be such a way that it blocked the laser beam to make system thinks that a wall was around the car.
There would be a sample of data reported by the laser periodically. For example, the car was so close to an object, but 50 msec later the car is at safe distance to objects around it. By confirming the laser data reported with weather forecast, system would mark this situation as rain fall or snow fall. Otherwise, system would report laser beam failed, and repair was suggested.
The laser beam on the side of a car would likely report intermittent safe distance and dangerously close repeatedly. However, to make system less expensive the photo sensors would be equipped at the front and back of the car, i.e. direction of a fast moving car.
By coupling a photo sensor and laser beam the car could keep safe distance with other objects.4. Clarification
By adding more laser beams or photo sensors in around a car would increase the chance to detect safe distance in snowy or rainy days, but it is not a bullet proof solution.
To detect a car was in its blind spot, the car could be equipped with more laser beams and photo sensors installed on the left side, right side, and corners of the car.
5. Explanation of devices
a. GPS navigator
With precision to lane, the control center could help to keep a car in correct lane. With knowledge of number of lanes on a road, control center could help the car to change lanes correctly.
b. Laser beam
The laser beam is using technology as used in radar. It beams waves and catches the rebound waves in order to determine rebounded signal strength. Based on the rebounded signal strength, it estimates the distance between the car and the object blocking the beams.
c. Photo sensor
The photo sensor would capture images. Image processing software would compare the captured image with its data in database for image recognition, i.e. snow, rain, car, walls, etc. in the photo. Based on the captured image, the control center would take appropriate actions.
I don’t think a photo sensor could estimate the distance between the car and object in the photo unless it knew the object’s size in advance. For example, photo of the back of a Toyota Camry is stored in database. Based on photo, system uses recorded real size of a Camry plus size of the Camry in the photo, it could estimate the distance.
d. Weather forecast center or car’s control center
Based on information received from a local Weather Forecast Station via wireless Internet, the control center could confirm and decide appropriate actions.
A photo sensor could also detect snow or rain based on sample data in its database stored in the control center.
The car’s computer system or control system could connect to the information control dashboard, where users could connect it to a cellular phone via WiFi.
6. Using photo sensors for traffic signs
Unless each traffic intersection implement signal transmitters to inform the current traffic light such as red, yellow, green AND stop sign, a car must use its photo sensors to detect traffic signs.
If traffic signal transmitters implemented, the car could catch the signals and adjust its speed, and use photo sensors to detect cars around it in order to move correctly.
The photo sensors could detect “green, yellow, red, or color arrow” or “stop sign” used in a traffic intersection. It compared to its database for a known patterns such as square box around a green, yellow, red spots.
Basically the car must have a database of images for known objects on the roads, so it can compare and decide the meaning of objects captured in a photo.
Basically the car must have a database of images for known objects on the roads, so it can compare and decide the meaning of objects captured in a photo.
7. Backup strategy in case of technology failed
Self-driving cars have been relying on first on GPS for lane precision, or second on lane marking to keep a vehicle travelling in correct lane.
However, military or GPS providers could take the satellites back for upgrade or different tasks. Thus the backup plan would be return to check the lane marking to guide vehicle correctly if GPS system failed. The lane marking is not a reliable source neither as many locations covered with snow during winter time. Snow plowing process would erase or fade lane marking, too. We cannot expect cities to maintain lane marking in perfect conditions.
If the system could not find solid lane marking for 1 meter or missed a broken lane marking, it should slow down; notify the driver to take over control; if the driver didn’t take over control for 2 seconds, it should turn on the hazard lights and bring the car into full stop on the current lane (the last known trajectory would be the best guess by the system in this situation) if it couldn’t change lane and park on the side road.
8. Wireless protocol
If many car manufacturers used the same laser beam and frequency for detecting safe distance between cars on the roads, then a simple protocol embedded in the laser waves should be implemented by those car manufacturers. For example,
- Common message ID: the first few characters to identify this is a laser beam message.
- Next 5 characters to identify a car manufacturer, e.g. TOYOT for Toyota
- The next 9 characters are proprietary to contain ID of the car, e.g. the first 2 characters for model, and the last 7 characters for unique ID of that car.
Since there wouldn’t be necessary for vehicle to vehicle communications to avoid hackers manipulate all cars on a road causing accidents, the last 9 characters reserved for each manufacturers providing flexibility.
Manufacturers should avoid using unique ID such as vehicle’s VIN and serial number as those numbers should remain private for law enforcement, insurance, and vehicle maintenance.
9. Calculating distance by using power transmitting and receiving from a car's antenna.
* Notes: The content of this section is copied from http://www.antenna-theory.com/basics/friis.php. By transmitting signals with known power and measure received power of signals bouncing back from an obstructing object or a car, we could calculate distance between our car and the obstructing object.
* Notes: The content of this section is copied from http://www.antenna-theory.com/basics/friis.php. By transmitting signals with known power and measure received power of signals bouncing back from an obstructing object or a car, we could calculate distance between our car and the obstructing object.
On this page, we introduce one of the most fundamental equations in antenna theory, the Friis Transmission Equation. The Friis Transmission Equation is used to calculate the power received from one antenna (with gain G1), when transmitted from another antenna (with gain G2), separated by a distance R, and operating at frequency f or wavelength lambda. This page is worth reading a couple times and should be fully understood.
Derivation of Friis Transmission Formula
To begin the derivation of the Friis Equation, consider two antennas in free space (no obstructions nearby) separated by a distance R:
Figure 1. Transmit (Tx) and Receive (Rx) Antennas separated by R.
Assume that PT Watts of total power are delivered to the transmit antenna. For the moment, assume that the transmit antenna is omnidirectional, lossless, and that the receive antenna is in the far field of the transmit antenna. Then the power density p (in Watts per square meter) of the plane wave incident on the receive antenna a distance R from the transmit antenna is given by:
If the transmit antenna has an antenna gain in the direction of the receive antenna given by GT, then the power density equation above becomes:
The gain term factors in the directionality and losses of a real antenna. Assume now that the receive antenna has an effective aperture given by AER. Then the power received by this antenna (PR) is given by:
Since the effective aperture for any antenna can also be expressed as:
The resulting received power can be written as:
[Equation 1]This is known as the Friis Transmission Formula. It relates the free space path loss, antenna gains and wavelength to the received and transmit powers. This is one of the fundamental equations in antenna theory, and should be remembered (as well as the derivation above).
Another useful form of the Friis Transmission Equation is given in Equation [2]. Since wavelength and frequency f are related by the speed of light c (see intro to frequency page), we have the Friis Transmission Formula in terms of frequency:
[Equation 2]Equation [2] shows that more power is lost at higher frequencies. This is a fundamental result of the Friis Transmission Equation. This means that for antennas with specified gains, the energy transfer will be highest at lower frequencies. The difference between the power received and the power transmitted is known as path loss. Said in a different way, Friis Transmission Equation says that the path loss is higher for higher frequencies.
The importance of this result from the Friis Transmission Formula cannot be overstated. This is why mobile phones generally operate at less than 2 GHz. There may be more frequency spectrum available at higher frequencies, but the associated path loss will not enable quality reception. As a further consequence of Friss Transmission Equation, suppose you are asked about 60 GHz antennas. Noting that this frequency is very high, you might state that the path loss will be too high for long range communication - and you are absolutely correct. At very high frequencies (60 GHz is sometimes referred to as the mm (millimeter wave) region), the path loss is very high, so only point-to-point communication is possible. This occurs when the receiver and transmitter are in the same room, and facing each other.
As a further corrollary of Friis Transmission Formula, do you think the mobile phone operators are happy about the new LTE (4G) band, that operates at 700MHz? The answer is yes: this is a lower frequency than antennas traditionally operate at, but from Equation [2], we note that the path loss will therefore be lower as well. Hence, they can "cover more ground" with this frequency spectrum, and a Verizon Wireless executive recently called this "high quality spectrum", precisely for this reason. Side Note: On the other hand, the cell phone makers will have to fit an antenna with a larger wavelength in a compact device (lower frequency = larger wavelength), so the antenna designer's job got a little more complicated!
Finally, if the antennas are not polarization matched, the above received power could be multiplied by the Polarization Loss Factor (PLF) to properly account for this mismatch. Equation [2] above can be altered to produce a generalized Friis Transmission Formula, which includes polarization mismatch:
[Equation 3]See also decibel math, which can greatly simplify the calculation of the Friis Transmission Equation.
-----------------------------------------------------------------------------
Notes:
The corner laser beam would also be used to estimate speed of the incoming car based on its distance for a period of time, e.g. closer after 1 second means that incoming vehicle is travelling at higher speed than your car. If the incoming vehicle is close and traveling faster, it's not safe to change lane at that time.
If blind spot/side is free, incoming vehicle is at safe distance and travelling slower or same speed, turn on change lane signal, and slowly change lane as expected.
* September 12, 2018: Governments shouldn't allow vehicle-to-vehicle protocol, because a hacker could cause malfunction in a car's sensor or system that leads to a chain accident.
A car should be equipped with a system to provide self-control of its operations.
* September 13, 2018: In the last accident by Tesla's autopilot, the car system alerted drivers to put hands on wheel or manually controlled the car 9 seconds before the crash happened, because the driver didn't follow autopilot's order.
Notes:
* April 1, 2018: This post only discussed "how to keep safe distance between a car and its surrounding objects." To change lane automatically would require more sensors and laser beams around the car to detect cars in left/right lanes.
The process to change lane safely would be similar to how a good driver does, e.g. checking blind spot, estimate other car's speed, etc.
* An example of using laser beam to detect incoming car in the lane that you wanted to move on.
The corner laser beam and side photo sensor would detect that there was no car running on the side and blind spot.
The process to change lane safely would be similar to how a good driver does, e.g. checking blind spot, estimate other car's speed, etc.
* An example of using laser beam to detect incoming car in the lane that you wanted to move on.
The corner laser beam and side photo sensor would detect that there was no car running on the side and blind spot.
The corner laser beam would also be used to estimate speed of the incoming car based on its distance for a period of time, e.g. closer after 1 second means that incoming vehicle is travelling at higher speed than your car. If the incoming vehicle is close and traveling faster, it's not safe to change lane at that time.
If blind spot/side is free, incoming vehicle is at safe distance and travelling slower or same speed, turn on change lane signal, and slowly change lane as expected.
* September 12, 2018: Governments shouldn't allow vehicle-to-vehicle protocol, because a hacker could cause malfunction in a car's sensor or system that leads to a chain accident.
A car should be equipped with a system to provide self-control of its operations.
* September 13, 2018: In the last accident by Tesla's autopilot, the car system alerted drivers to put hands on wheel or manually controlled the car 9 seconds before the crash happened, because the driver didn't follow autopilot's order.
What Tesla was supposed to do in the case driver didn't put their
hands on the steering wheel as alerted?
- Turn on hazard lights
- Slow down the vehicle and then stop
- If it was a self-driving car, try to slow down, change lane, and then park on the side way.
Tesla
didn’t test or implement solution for this case.
No comments:
Post a Comment