RADAR vs. LIDAR sensors in automotive industry
As we are approaching a future of self-driving vehicles, the number of sensors in vehicles will increase. Currently, driver assistance systems use a combination following key elements:
- RADAR sensors (Radio Detection And Ranging)
- LIDAR sensors (Light Detection And Ranging)
- Cameras
- Ultrasonic
Focusing on 2 sensor technologies, RADAR and LIDAR, they both have physical limitations and often overlapping capabilities. An answer to a question which sensor has a domination position in autonomous car industry remains ambiguous.
RADAR is short for Radio Detection and Ranging and is considered as a conventional system which has been used for many in industries such as air traffic control, air-defense systems, antimissile systems, aircraft anti-collision systems, ocean surveillance systems, radar astronomy, space surveillance, meteorological precipitation monitoring and automotive industry for a couple of years. RADAR is well established mainly in automotive industry. Using radio waves, a radar has a capability to determine an angle of objects, velocity, and range. A RADAR categorization is based on different operating distance ranges, starting from 0.2 m to more than 200m.
Then, we recognize the following radar sensors:
- Short range radar (SRR)
- Medium range radar (MRR)
- Long range radar (LRR)
Advantages of RADAR is a lighter weight, a capability to operate in different conditions in comparision with a LIDAR and a longer range. However, all RADAR categories have a limited field of vision. A main limitation of the LRR is not enough accurate reaction to certain events, f.e. a car cutting in front of the vehicle or a distance setting due to the road curvature. For instance, small vehicles, such as bicycles and motorcycles, or any other vehicles not driving in the center of the lane remain undetected by a RADAR sensor. Also, any RADAR normally loses a sight of the target vehicle in curves and struggle and may not detect a vehicle at close quarter. In addition, if multiple objects are close to each other, the RADAR sensor may get confused by them as being one large object and return one centroid range. Then the sensor generates a lot less data than is needed. To reduce an inaccuracy of the sensor, an automotive industry pairs a RADAR sensor and camera together to enhance a more precise detection system and emergency breaking.
LIDAR is an acronym for Light Detection and Ranging uses. Currently, LIDAR is one of the most popular sensors on the market able to scan more than 100 meters in all directions and generate a precise a 360° 3D view of the surroundings and obstacles that a vehicle should avoid. LIDAR sensor is also more target selective than any radar. LIDAR normally measures a distance to an object by calculating the time taken by a pulse of light to travel to an object and back to the sensor. A data collected is then used by car to make intelligent decisions. LIDAR in automotive systems typically use 905nm wavelength that can provide up to 200m range in restricted FOVs and some companies are now marketing 1550nm LiDAR with longer range and higher accuracy.
However, to compare LIDAR with RADAR, LIDAR systems are larger and more expensive, usually mounted outside of the vehicle. Smaller and cheaper lidars are being developed and one of the most promising seems the miniature form of a silicon chip. Again, LIDAR has great capabilities in light conditions but starts to struggle with snow, fog, rain, and dust particles in the air due to its use of light spectrum waves. Also, LIDAR is not able to detect a colour, or contrast, and provides a poor optical recognition,
How a car with lidar sees the world.
Regardless of whether LIDAR is better or not, no sensor solution is enough viable yet. Each sensor type has its advantages and disadvantages and the amalgamation of multiple sensors and cameras is required in order to get the highest accuracy possible.
References
1 https://www.technologyreview.com/s/603885/autonomous-cars-lidar-sensors/
2 https://blog.nxp.com/automotive/radar-camera-and-lidar-for-autonomous-cars
3 https://cleantechnica.com/2016/07/29/tesla-google-disagree-lidar-right/
4 comments on “RADAR vs. LIDAR sensors in automotive industry”
Comments are closed.
She also worked as a project associate in European Strategy Council, working primarily on the Innovation and Cybersecurity Forum.
Michaela holds a master degree in International Relations & International Law from the University of Roehampton in London, Certificate in Technology progress from EUIC and Diploma of International Affairs and Politics from the University of Coast Opal.
Michaela also lead a research project on the ICT technology with Crucible Research Center.
Hi Michaela, I enjoyed your paper. I often read about Google’s Waymo project. According to a recent article from Bloomberg [1], the company has spent a great deal of time working on “scalability.” The use of LIDAR has apparently lead to a 90 % decrease in the cost of the LIDAR sensor, which was considered to be the most expensive item in a self-driving car solution. I found a great article on the use of LIDAR for Googles WAYMO project.
The first public Google Self-Driving Car prototype was built on a Toyota Prius. At the time, it used the Velodyne HDL-64E LIDAR sensor, which cost about $75k. The LIDAR sensor needed to be up high to see around the vehicle, so Google mounted it on a large riser. I remember these cars driving around Mountain View when I lived in the Bay Area. The LIDAR sensor would spin 360 degrees so that it could detect the environment, but wasn’t very good at seeing things close up. not great at detecting up-close. To fix this, Google augmented the LIDAR input with several radar sensors all around the vehicle to compliment the roof mounted LIDAR.
After doing a byte of research, and reading your Blog, it looks like Google has made the Waymo project much more scalable and cost efficient, not to mention, the vehicles utilize Serverless-compute and the AI exists within the vehicle making it a true autonomous system. Reducing the costs will be a major factor in the adoption of self-driving vehicles.
John Krafcik, Waymo’s chief executive officer, told Bloomberg “We’ve made tremendous progress in our software, and we’re focused on making our hardware reliable and scalable.
This has been one of the biggest areas of focus on our team for the past 12 months.” Krafcik also told Bloomberg the new sensor package on the Waymo Chrysler Pacifica is “highly effective in rain, fog, and snow,” which have typically been trouble for LIDAR systems thanks to the reflective nature of water in the air. [2]
[1] https://www.bloomberg.com/news/articles/2017-01-08/alphabet-s-waymo-cuts-cost-of-key-self-driving-sensor-by-90
[2] (https://arstechnica.com/cars/2017/01/googles-waymo-invests-in-lidar-technology-cuts-costs-by-90-percent/_
Very insightful post. In my opinion, LIDAR sensors are an integral part of the autonomous drive system. Recently, Audi released the world’s first production level 3 autonomy car, its flagship A8. In addition to the array of cameras and radars, the A8 is also the first production car to use LIDAR. As a result, the car can detect smaller objects such as bicycles, poles and debris on the road, thanks to the precision of LIDAR. The benefit for the user is that the A8 does not need the driver to pay any attention to road situations up to 60km/h. Nvidia is is powering Audi’s autonomous drive system, processing data at 2.5 billion inputs per second. I believe as autonomous drive technology progresses, LIDAR is going to be standard equipment on any premium car, both for safety and convenience.
Michael Zhang
MS&E 238A student
Nice post! One question I had is around the cost of LIDAR vs RADAR solutions; from some light reading I noticed that LIDAR chips (made by companies like Analog Devices) are quite expensive and show no immediate signs of reducing despite higher production volumes (since production costs are high). Curious if you came across anything for/against this.
Hey Guarav. Thanks for your comment. As you said, a price of the LIDAR is significantly higher than a radar sensing technology. This is one of the reasons why companies pursuing a creation of autonomous car excluding LIDAR in order to maintain a customer friendly market price of such a car.
An inclusion of the LIDAR can add almost $8000. Considering downsides of the LIDAR technology, it is not surprising there is a slow approach when investing into the LIDAR.