We All Want to Have an Autonomous Car; Here’s How They Work.

Autonomous driving will gradually come to all cars to be developed in the future, especially electric vehicles, while legislation evolves to allow drivers to be utterly unconcerned with the controls. Naturally, this situation concerns other road users such as pedestrians, cyclists, and motorcyclists. In particular, associations of the last doubt that autonomous cars are capable of detecting them in all circumstances; however, some of the most common problems faced by this type of vehicle have to do with the timely detection of the distances separating it from other objects, particularly smaller ones such as motorcycles, bicycles, pedestrians, and animals that may be encountered on the road.

One of the best options is lidar sensors in the race to improve sensors and their timely detection of obstacles. Lidar is an acronym (Light Amplification by Stimulated Emission of Radiation), translated as a system for measuring and detecting objects using lasers. 

Lidar sensors are one of the elements that enable these vehicles to recognize the environment as if they were the driver’s eyes. Laser pulses emitted from a lidar system are reflected from objects on and above the ground surface: vegetation, buildings, bridges, and so on. An emitted laser pulse can return to the lidar sensor as one or many returns. Any emitted laser pulse that encounters several reflecting surfaces as it travels toward the ground is divided into many returns as there are reflecting surfaces.

The first laser pulse returned is the most important and will be associated with the largest entity in the panorama, such as a treetop or the top of a building. However, the first return may also represent the ground, in which case the lidar system will only detect one return.

Multiple returns can detect the elevations of various objects within the laser footprint of an outgoing laser pulse. Intermediate returns are used for vegetation structure and the last return for bare ground terrain models.

The last return will not always be from a ground return. For example, consider a case where a pulse hits a thick branch on its way to the ground, and the vibration does not reach the ground. In this case, the final return is not from the ground but from the branch that reflected the entire laser pulse.

Additional information is stored along with each x, y, and z positional value. In addition, the following lidar point attributes are maintained for each recorded laser pulse: intensity, return number, several returns, point classification values, points that are on the edge of the flight line, RGB values (red, green, and blue), GPS time, scan angle, and the scan direction.

This requires the infrared laser beam emitter, a lens that collects the light beams as they bounce back, and a chip or system that processes all this data to build a 3D map of the scene in front of the sensor. 

There are two types of sensors, fixed and rotating or 360-degree sensors. For example, the fixed ones, when used for driving, are placed at the top of the windshield or on the sides of the vehicle to achieve a broader spectrum of vision and sometimes integrate a camera that records the lines of the road and traffic signs.  Meanwhile, gyroscopes are mainly used in topography to study the relief of terrain. However, they have also been adapted for cars, and you can see them on the roof of autonomous car prototypes.