Autonomous vehicles have recently been receiving lots of attention, especially due to research advancements by large corporations such as Google. It is unlikely that these vehicles will be used on a widespread capacity in the next few years. However, there are some applications for which this technology is already available.
Figure 1: Google has been doing extensive road tests of autonomous vehicles, both as a) custom vehicles and b) modified standard cars.
Autonomous vehicles require various technologies, such as sensors, actuators, algorithms, and processors.
The sensors and actuators are divided into subcategories:
- Navigation and guidance: ensuring you get to where you want to go
- Driving and safety: ensuring that the vehicle remains safe
- Performance: ensuring the car is running optimally
Navigation and Guidance
For an autonomous vehicle to function, they need to know where they are, where they are going and the best way to get there. Traditionally, this has been carried out manually through many different methods, such as the compass, sextant, LORAN radiolocation, and dead reckoning.
However, these methods all have varying degrees of accuracy, consistency, and availability. In a self-governing vehicle, these factors must be consistently checked to ensure that the vehicle is traveling to the correct location. This would, for example, ensure that unexpected diversions do not prevent the vehicle from reaching its destination.
The main system which is used is the Global Positioning System (GPS) receiver. This measures and assesses the real-time position of the vehicle through various signals from satellites. Overall, this produces a very accurate measurement of location (Accurate to 1 meter) but requires 30-60 seconds in order to obtain a reading.
Recent developments now mean that GPS subsystems are available on system on a chip (SoC) IC or multi-chip chipsets. These technologies only require power via an antenna and can perform lots of complex calculations. Lots of these ICs have a centralized RF preamp for the 1.5-GHz GPS signal. However, these autonomous vehicles often place the antenna and a low-noise amplifier (LNA) RF preamplifier upon the roof.
Figure 2: A complete GPS module, such as this F4 unit from Linx Technologies, requires only an antenna and CD power to provide position data via a serial interface and in standard format to a system processor and its mapping application.
The polarization of the GPS signal must be matched by the right-hand circular polarization characterization (RHCP) of the antenna. Furthermore, the antenna can be either a ceramic-chip unit or a small wound stub form.
GPS is a fundamental technology for self-governing vehicles. However, if this signal is prevented by structures such as tunnels, or is hindered by radio interference, then this loss of signal must be accounted for. This is carried out by an inertial measurement unit (IMU) which is placed upon the vehicle. Overall, 3 gyroscopes and 3 accelerometers are involved, with a pair for each axis. Both the rotational and linear motion is monitored which can then be translated to calculate the movement of the vehicle.
Without MEMS-based gyros and accelerometers, IMUs are not as practical as it is too big, expensive and requires too much power.
Figure 3: MEMS devices have radically changed the implementation of IMU functions such as gyroscopes and accelerometers; this tiny IC from STMicroelectronics incorporates a trio of orthogonal gyroscopes and provides a digitized, serial output of their angular readings.
Driving and Safety
Another requirement for autonomous vehicles is the necessity to be able to observe what is happening around them. This could be achieved through a variety of cameras, however, the mechanical and processing issues associated with cameras make them ineffective.
Additionally, the lack of depth perception and lighting issues also make them inaccurate. Therefore, the most commonly used mechanism of vision is a LIDAR system (light detection and ranging), which is a 360 degree rotating, scanning mirror assembly. This establishes 3D information and allows the vehicle to avoid collisions.
Through the applications of LIDAR, fast and powerful laser-light pulses calculate the distance of an object through reflected light calculations. Timed cameras obtain and improve the relative resolution of this data which can be used to calculate the position, speed, and direction of the car in relation to objects outside of the car. This information can then be used to avoid crashes and instruct the car to change direction if required.
This LIDAR system, however, is less effective in the control of close interactions. For example, parking, changing lanes or whilst in traffic. Radars, however, can supplement their function to ensure accuracy. These radars are built into the sides of the cars in an integrated design, operating at the 77 GHz frequency and demonstrating significant resolution.
Several other components are required, such as AD8283. This involves an antialiasing filter (AAF) with a one direct-to-ADC channel, a single 12-bit analog-to-digital converter (ADC), 6 channels of a low noise preamplifier (LNA) and, a programmable gain amplifier (PGA).
The key use for the AD8283 is in the fast ramp, frequency controlled continuous wave radar (HSR-FMCW radar). Optimization of each block is used to ensure that the demands of the radar system are met, ensuring that PGA gain range, LNA noise, AAF cutoff characteristics, and ADC sample rate and resolution are equalized.
This specific radar has a multiplexer which alters between these channels once each sample is acquired. These channels have a gain range of 16 to 34 dB in 6 dB increments and a maximal conversion rate of 72 MSPS. Effective performance is ensured by a combined input-referred noise voltage at a maximum gain of 3.5 nV/√Hz.
Figure 4a: Radar System overview of the AD8283, a 6-channel radar receive path AFE.
Figure 4b: A simplified block diagram of a single channel of the AD8283. Automotive radar systems require a) sophisticated, controllable analog front end circuitry to handle the reflected pulse signals across multiple receiver channels; b) the AD8283 from Analog Devices is designed specifically for this situation.
The performance of autonomous cars is a vital component, although it may not be as glamorous as the navigation and guidance aspects. This involves various processes, such as the monitoring of power, consumption and thermal dissipation.
The current and voltage of batteries also need to be monitored through isolated sensing. Low-voltage circuit boards alternatively use a high-side, current-sense, milliohm resistor (shunt) in combination with an amplifier which monitors the voltage decrease across it.
Although the amplifier is used with a discrete sense resistor, there is now an alternative that saves space, minimizes errors in readings which are primarily due to thermal drift of the sense resistor as it self-heats, and simplifies the bill of materials (BOM) by reducing the number of parts.
The INA250 from Texas Instruments puts a sense resistor and differential amplifier in a single package, thus resulting in a far-smaller board-layout footprint, fewer circuit-layout problems, and lower system cost due to simplified schematic, Figure 5.
Figure 5: Simplified schematic of the INA250 from Texas Instruments. Measuring current and thus power is a vital housekeeping function needed in most circuits; the INA250 current-sense resistor plus differential amplifier components eases design and PC-board layout tradeoffs while guaranteeing high precision and accuracy along with lower cost.
The self-driving vehicle is the subject of great discussion. However, it is not yet known when, or if, these cars will be used in mainstream applications. However, further development of complex algorithms and processors is required.
This information has been sourced, reviewed and adapted from materials provided by Mouser Electronics.
For more information on this source, please visit Mouser Electronics.