Autonomous Vehicle

Connected Autonomous Vehicles: the journey and the road ahead

Connected Autonomous Vehicles (CAV) is a radical technology that has great potential to improve the efficiency of public transportation systems and reduce vehicular accidents.

The transport industry is in transition; like every other sector it is driving towards automation. Leading car manufacturers like McLaren and Ford are leading the way in researching consumer needs and the business model required to fully embrace autonomous technology.

As per a report from the consultancy firm Boston Group, connected automated vehicles would have a market share of 53% by 2025, which will grow exponentially to 77% by 2030.

The Evolution of Autonomous Vehicles

When Google introduced autonomous vehicles in 2017, the original price tag was $80,000. Today, the price is approximately one-tenth of that.

Though driverless cars may still be a distant dream for many, the investor interest has not diminished for the technology.

Autonomous Vehicle Implementation Prediction by Victoria Transport Policy Institute, December 2015

However, the history of driverless cars can be actually traced back to the 16th century with Da Vinci’s self-driven cart. In this model, high tension springs provided the power output while the steering could be set in advance for a predetermined path. This ingenious design is often regarded as the first robot in this world.

When Robert Whitehead invented a torpedo that could travel several miles underwater while maintaining depth, it proved to be a revolutionary step forward for guidance systems with numerous weapons and other autonomous devices being designed with inspiration from Whitehead’s tropedo guidance mechanism.

But perhaps the greatest step towards autonomous technology was the first autopilot aircraft designed by Mechanical Mike. Gyroscopes, which are an essential part of autonomous vehicles today, kept track of the autopilot plane’s control for following an accurate direction.

Subsequent developments in automation technology were made with the invention of cruise control, micro-processing modules like cameras and the idea of remote-controlled lunar landings.

The US Department of Defence Research conducted a series of experiments which further improved autonomous technologies drastically. In 2007, four self-driven cars navigated 60 miles in an urban environment within the allotted 6 hours’ time limit.

And when Tesla introduced its autopilot feature in 2015 which enabled hands-free control for highway driving, the autonomous technology was promulgated to the world as a revolutionary step forward towards safe driving and improved transport systems.

Recent advancements in CAV technology

Recent developments in CAV technology have taken leaps in ensuring passenger safety self-driven vehicles.

Car-makers and tech companies have joined hands to develop the technology with a microscopic focus on crash notifications, prior accident warning, over-speeding, presence of fog or black-ice, and so on. Famous tie-ups include Microsoft partnering with Volvo and Toyota for the development of autonomous cars, while Apple and Uber have joined hands on autonomous car projects.

In Europe, BMW and Mercedes-Benz are already leading the way with several autonomous prototypes currently in the designing phase.

Recent developments in the technology include the following:

  • With 20 million miles clocked by the driverless cars of Waymo, an autonomous car technology company founded in 2009, the company is regarded as the leader in CAV technology. The company is closest to reaching the level 5 of automated driving with its self-driving systems capable to operate without any human interference. Waymo currently stands at Level 4 of automated driving which can completely stop a vehicle in case of any system failure. The image below shows how Waymo technology operates:
Beaming a million Laser pulses per second, the LIDAR system measures the reflection return time to evaluate its surroundings. It uses wavelengths to track objects and motion. Source: https://www.researchgate.net/figure/Autonomous-car-Waymo-and-its-instrumentation-system_fig1_326900783/amp

In addition to the LIDAR technology, Waymo also uses GPS sensors to detect audio sirens from emergency vehicles.

  • Cruise, another autonomous car company founded in 2013 released the Origin, a fully driverless car without a steering wheel or pedals. Designed for highway speeds, the vehicle has a lifespan of 1 million miles, which is about 5 times more than the lifespan of an average car. Origin is due to be used under Cruise’s ride sharing services as a taxi. The vehicle has ‘owl’, a hybrid sensor assembly using both the radar and camera.
Cruise Origin Interior Design. The digital screens overhead displays passenger and route information. Source: https://www.slashgear.com/cruise-origin-electric-autononous-shared-transport-gm-honda-revealed-21607372/amp/

  • Argo AI is another company working towards a fully integrated self-driving system. The company is working with Ford and Volkswagen to develop Level 4 of automation technology. Argo AI seeks to integrate LIDAR, radar, cameras and light detection sensors to develop a fully automated vehicle. Argo is using machine learning algorithms to solve the technical challenges of perception and decision-making in automated technology. In 2019, Argo AI partnered with Carnegie Mellon University for $15 million research to study advanced algorithms for autonomous vehicles. The company uses alternate systems for the driver to take manual control in the event of a power loss along with a power backup system for braking, steering and sensor components.

Although the world is rapidly progressing towards an automated driving future, the application of the technology has been limited in India to cruise control and development of v2v systems.

This is due to the lack of in-depth driving data and poor infrastructure coupled with several basic challenges like lack of high-speed internet service in many areas.

However, this has not dampened the spirits of IT firms and autonomous vehicle start-ups to try and navigate these challenges for a better automated future.

  • Ati Motors, a Bengaluru-based start-up has started developing autonomous electric cargo vehicles with features such as electro-mechanical braking, OTA software updates, lithium-ion battery, etc.
  • Flux Auto is another Bangalore based self-driving tech company for commercial vehicles. The company is currently developing its software which can be retrofitted in trucks with cruise control, lane keeping and anti-collision features. Flux Auto uses cameras instead of LIDARs to drive down costs.
  • Swaayatt Robots is a developer of self-driving technology for on and off-road vehicles. The Bhopal-based start-up has claimed its motion, perception and decision frameworks can handle the complex traffic situation on Indian roads.

Roadblocks in CAV technologies

In order to ensure development and commercialization of automated cars, there is a need to navigate through several tech and other challenges that lie ahead:

  • Safety

The reliability of self-driven cars would depend on the distance covered by the car. As per the requirements, the autonomous car should be able to drive 291 million miles without any accident for a 95% equivalence to human drivers.

Another major concern is testing the vehicles in ‘difficult miles’, or challenging road situations like diversions, constructions, etc. So far, prototypes have been tested in ‘easy miles’.

In 2018, the first fatal accident occurred when a Level-4 Uber prototype hit a pedestrian who was crossing the road. As per the NTSB report, the pedestrian was detected, but not identified correctly.

The University of Michigan proposed its own autonomous car testing facility ‘Mcity’ which tests the performance of automated vehicles based on accelerated evaluation, behavior competence and corner cases.

Accelerated evaluation focuses on collecting driving data pertaining to lane changing, left turns and car following.

Under behavior competence, the driver’s behavior is made risky to focus on navigating through ‘difficult miles’.

The test results were then evaluated and 16 safety scenarios were identified.

  • Testing and Validation

Autonomous driving systems require a comprehensive testing framework, as the software is too complex and any decision made by the system would affect human lives directly.

Traditional validation and testing measures are not feasible for autonomous systems due to the high level of uncertainty involved.

Alternate approaches like machine learning may be able to design a more comprehensive system for advanced decision making. But for processes like object detection, the algorithms classifiers need to be subjected to a large amount of data, making the process much more complex.

The Fail-system design is another challenge, as two independent subsystems are required to ensure that one takes over if the other one fails.

  • Object Identification in relation to its surroundings

Dynamic situations on the road like diversions, construction sites, missing road signs, etc. are contributing factors towards orientation challenges on road.

Measures such as real-time image processing and machine learning integrated in Tesla cars enables the software to adapt to dynamic road conditions.

Other companies rely on LIDAR for a 3-dimensional map of the surroundings.

Creation of smarter environments is yet another approach which can reduce the complexity of such challenges by informing the vehicle of any change in the surroundings.

  • Legal challenges

The need of a regulatory framework is pivotal for the development and deployment of CAV technologies. With driverless cars, there is a shift of liability from drivers to car makers regarding accountability in case of any mishaps.

All major consumer concerns have to be considered before actual deployment of Level 5 automation technologies on road.

  • Financial Challenges

The development and deployment of autonomous vehicles is a steep challenge financially. The software and hardware components required for CAV technologies like sensors and communication devices are considerable, which may limit Level 5 automation features to premium tier models exclusively. This would directly affect the end consumer regarding affordability.

Ride-sharing models like Cruise origin could be the answer during the first phase of mass deployment of CAVs.

However, it is imperative to make the technology affordable in the near future to ensure mass adoption of autonomous vehicles.

  • Ethical challenges

As per a 2017 report by the German Ethics Commission, ” In the event of unavoidable accident situations, any distinction based on personal features (age, gender, physical/mental constitution) is strictly prohibited. It is also prohibited to offset victims against one another. General programming to reduce the number of personal injuries may be justifiable”.

But, due to the immense complexity of on-road situations, autonomous cars may face situations where they need to make a choice between putting the lives of the passengers at risk or saving a pedestrian. Even with the presence of regulatory frameworks, this can be a daunting task due to the moral characteristics of individuals.

The future of autonomous cars depends on overcoming this challenge, which can only be done with the consent of all stakeholders after thorough analysis on the benefit and risk of each decision made in this regard.

Towards an autonomous future

Although there are many roadblocks and challenges for implementation of CAV systems, the benefits that they may bring is far greater than the risks involved. The key to overcome all the challenges and limitations is to have the best solutions, which starts from choosing the best hardware. Working with the best IoT devices manufactures I have realized that there is nothing like cheap and best, at the same time there can be a huge difference in accuracy and performance in the best and the second best. So always go for highest rated best performing hardware start with having a proof of concept, test these devices and build your IoT ecosystem on these pillars.

If successful, CAV has the potential to drastically improve transportation systems, driver safety, fuel and parking efficiency, emission reduction, etc.

In a developing country like India, where public transportations are already stretched to the limit by catering to a population of 1.38 billion, CAV technology is definitely the future.

For successful implementation of CAV technologies, the need is a conglomeration of all stakeholders, from manufacturers, tech companies and regulators to arrive at a general consensus on how to best use this technology and limit the risks involved in autonomous vehicles.


“With great power comes great responsibility”, we at Teltonika do realize this since decades and we have been making sure we do our bit to help IoT ecosystem grow. Our teams offer special attention to new players in the industry and help them wherever needed with their expertise, over the time we have provided our constant support to many Integrators and they have grown exponentially over the time.

Author:

Praveen Dhusia

Business Development Manager

Teltonika India

Indian Institute of Management Alumnus with 8+ Years of Experience. I hold a degree in Electrical and Electronics Engineering and have expertise in IoT solutions. Worked on some landmark projects like Buddha International F1 Circuit, Redevelopment of Connaught place, IGI Airport Terminal 3 (MEP) etc. In my free time, I along with my friends try to give back to society by helping underprivileged young kids with their education.

Published in Telematics Wire

Related Articles

Back to top button