ADAS – The technology differentiator in the automotive Industry

241

“90% of the innovations in a car are in the software. 100% of cars will connect to the Cloud. The car is no longer the stronghold of Ford, Daimler, Toyota or Nissan but rather an opportunity for all technology companies and solution providers, from Tesla to Google, from Uber to Lyft, from Apple to Baidu, and also FPT”, said Nguyen Duc Kinh, director of FPT Software Global Automotive (FGA).

Advanced Driver Assistance Systems (ADAS)

One of the most fascinating abilities of intelligent beings is the seamless perception  and  interaction with the environment. Guidance and control of automobiles make up a comprehensive example for these abilities. A human driver needs to perceive and understand the automobile’s environment.  Based on the understanding  of  their  surroundings,  the  driver  can  plan,  initiate,  supervise and control the automobile’s behavior. Driver assistance systems  aim  to  replicate  those  capabilities on artificial systems.

ADAS are systems developed to help the driver for safer and  better  driving.  ADAS  offer  many  important  research problems to work on such as the development of different types of sensors, the processing of sensor information, the detection and tracking of pedestrians and vehicles, behavior and intent analysis of drivers as well as human factor interfaces.

Currently  demand  for  the  development  of  ADAS  is  greater  than  ever  due  to  the  need  for  safer  vehicles  and roads. Since ADAS and automatic vehicles require more  complex  sensory  input,  the  automotive  industry  is  facing  a  big  challenge  in  bringing  autonomous  vehicles  to  life.  ADAS  requires  much  more  complex  integration   than   other   sensor-based   technologies;   a  well-developed

ADAS  needs  a  wide  application  of  data integration techniques.

ADAS – The technology to make a difference FPT has invested in the research and development of ADAS technology for years. FPT’s research focused on problems around the relationship between vehicle real-time detection, object tracking, lane detection and car-to-car distance estimation.

The project to estimate the distance between vehicles for safe warning achieved over 93% accuracy. Several accident alert solutions by monitoring the driver’s status also received many improvements. There is also a pedestrian detection solution to help alert drivers to sudden movement on the road. The  number  plate  recognition  (NPR)  algorithm  was  developed as an important feature for an ADAS system. By using multiple object detectors such as Haar-like,  Local  Binary  Pattern, Deformable  Part  Model for both detection and classification, the system can detect cars and motorbikes in heavy traffic conditions with an accuracy of 95% during the day and 85% at nighttime.

The  NPR  system  can  handle  Vietnamese  and Japanese license plates and remains stable under different camera positions and weather conditions. These  successes  form  the  basis  for  FPT  to  move  deeper  into  the  automotive  sector  when  this  industry is experiencing a significant transformation by the impact of technology trends like Cloud, Internet, telecommunication infrastructure, IoT etc. FPT Software and big opportunities in the automotive industry.

Nowadays,  with  the  development  of  electric  vehicles  and  self-driving  cars,  the  role  of  software  in  the  automotive  industry  is  becoming  more  and  more  important.  Typical  software  applications  in  the  car  include:  navigation,  infotainment  software  (media/audio/video),  display  software  (electronic  speedometer,  light  symbols, air conditioning), and control software for the ECU (Electronic Control Unit – the embedded system controlling one or more electrical systems of the vehicle). In modern automobile design, software is considered the brain which ensures the operation of the car.

Almost  all  FPT  customers  in  the  automotive  industry  are  Tier  1  and  Tier  2  companies  such  as  Panasonic,  LG  Electronics,  Denso,  Continental,  NXP  or  Renesas.  Sixty percent of these customers are from Japan, 25% are from the US, and the rest are from Germany, Korea and other countries. In 2017 FPT will focus on building R&D capabilities related to image processing and recognition, to be applied in Advanced Driver Assistance Systems (ADAS) to better satisfy the customer’s needs.

The Director of FGA pointed out that affordability, enthusiasm and youthfulness are advantages of FPT Software personnel compared to international competitors. FPT is in possession of ADAS technology as well as many important components to create an Internet of Vehicles (IoV) platform, including hardware (Head-Unit equipment) attached on vehicles and Cloud, and could look to create big breakthroughs in the automotive market.

Nguyen Duc Kinh, Tran Nguyen Ngoc – FPT Software
(Published on the FPT Technology Magazine, FPT TechInsight No.1)

Related posts: