Techday
Physical AI HBM Smart Factory SDV AIoT Power Semicon 특수 가스 정정·반론보도 모음 e4ds plus

[Product of the Month] Door-cock prevention solution using mmWave sensors!

Google 우선 소스 기사입력2019.07.08 10:45

- 140' azimuth secured
- Stable against noise interference
- Minimize adverse environmental conditions such as snow and rain


You have probably had the experience of opening your car door and not noticing the door or pillar of the car next to you, or of opening your car trunk automatically and hitting the roof of your garage. In such cases, you will end up in an unfortunate situation where your car is damaged and you will have to pay a high repair bill.

Car manufacturers have looked into a variety of sensing solutions in the past to address the above issues. Most sensors have limitations in reliably detecting objects located in all directions around the car under various environmental conditions or locations. For example, some sensors only detect objects in one direction.

In comparison, radar sensors have the advantage of being able to “see” a wide 3D space around the sensor. In particular, millimeter wave (mmWave) sensors are installed in a size small enough to be invisible inside the vehicle door, so they do not affect the vehicle design, and they perform relatively stably even in noise interference such as ultrasonic sensors within the sound wave frequency range.
[Figure 1] A millimeter wave sensor on the vehicle door detects an external obstacle.
[Figure 2] The millimeter wave sensor on the lift gate detects the garage roof.

Cameras, which are typical vision sensing equipment, have limitations in that they have difficulty detecting objects in dark places or when the lens is covered with snow, dust, or mud. However, millimeter wave sensors do not require lenses, so they can accurately detect obstacles around vehicle doors and trunks even in a variety of environmental conditions, such as snow, rain, and glare.

Additionally, the millimeter wave sensor acts as a side impact radar along with the smart door opening function, and can detect other cars, pedestrians, or stationary obstacles around the car while it is moving.

“TI’s AWR1642 sensor has a multi-mode programmable digital signal processor (DSP) that can be configured to detect objects in a 3D space up to 1-2 meters around the car’s doors or trunk when parking, considering smart door opening, and the sensor can also operate while driving to detect side impacts,” says Kishore Ramaiah, product manager for automotive radar at TI. “By using a millimeter wave sensor, it can also avoid adverse vision conditions.”

In addition, the AWR1642 millimeter wave sensor can be used for functions such as curb and curb height detection, parking assistance, and when the sensor is mounted on the front of the vehicle, it can even recognize obstacles such as speed bumps and potholes in the road.

Obstacle detection
Below is a demo that demonstrates how accurately TI's AWR1642 millimeter wave sensor can detect objects such as a plastic pipe, a wooden board, a metal object, and a traffic cone at various distances and fields of view. This is a scene of an experiment where an obstacle detection sensor was placed on a tripod 0.5 m above the ground, and the detection ability was observed after placing objects at a distance of 0.5 m and 1 m from various angles.
[Figure 3] The obstacle detection sensor is detecting a wooden board in the experimental environment.

Experimental results showed that most obstacles could be accurately detected up to ±70° azimuth.

Curb height detection
[Figure 4] is an experimental environment for detecting the height of a curb placed as an obstacle. A 19 cm high curb was placed 30 cm away from the sensor, and the sensor was placed at a height of 90 cm from the ground, similar to the height of a door handle of a typical car.

[Figure 5] shows the range profiles measured by the sensor, one is a reflection from the curb edge and the other is a reflection from the corner formed by the ground-curb contact. By measuring the angles corresponding to these reflections, the curb height can be estimated. The maximum height estimated through this experiment was 20.5 cm, which is close to the height of the curb actually placed as an obstacle, which is 19 cm.

This information could be used for automatic door alerts, which alert drivers and passengers to potential obstacles outside the car when the doors are opened.
[Figure 4] Obstacle detection sensor evaluation module placed in front of a 19cm high curb, sensor distance is 30cm
[Figure 5] The detection points of the curb, the curb edge, and the ground-curb contact point can be identified as processed data.

The obstacle detection sensor evaluation module (EVM) in [Figure 6] uses the AWR1642 sensor, which implements a wide field of view even at close range through a single patch antenna. It also has the advantage of being able to detect both azimuth and elevation directions, allowing it to identify obstacles within 3D space.
[Figure 6] Obstacle detection sensor evaluation module based on AWR1642 sensor
[Table 1] AWR1642BOOST-ODS Evaluation Module Parameters

The above tests show that sensors placed in car doors, trunks and around the car can reliably help drivers detect and avoid obstacles in parking spaces and on the road.

This article is a summary of an article titled, “How to Prevent Door Cocking in Cars with Millimeter Wave (mmWave) Sensors” by Kishore Ramaiah, Product Manager for Automotive Radar at Texas Instruments.