Are Data Collection Concerns Delaying Self-Driving Cars?
By John Oncea, Editor
While data concerns aren't the only reason self-driving cars are taking longer to become mainstream, they are certainly a significant factor.
There’s been much written about the current and future status of autonomous vehicles (AVs). I tried to put an end to several of the myths surrounding them, linking to four other articles I wrote about the topic in the past year and a half. Unfortunately, there is no definitive answer as to when – if at all – self-driving cars will be a reality.
Rather than kick that topic around again let’s do something different and dig into one specific obstacle – data collection – playing a role in the arrival of commercially available AVs. And to be clear, we’re talking about Level 5 autonomy: fully self-driving vehicles requiring no human intervention.
What Is A Level 5 Autonomous Vehicle?
Since we’re specifically looking at Level 5 autonomy let’s agree as to what that means.
SAE International launched SAE J3016 Recommended Practice: Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles, commonly referenced as the SAE Levels of Driving Automation, in 2014. Since then, it has served as the industry’s most cited source for driving automation.
“With a taxonomy for SAE’s six levels of driving automation, SAE J3016 defines the SAE Levels from Level 0 (no driving automation) to Level 5 (full driving automation) in the context of motor vehicles and their operation on roadways,” SAE writes.*
Following is a breakdown of the levels and what’s currently available:
- Level 0 (No Automation): Most cars on the road today fall into this category, where the driver controls everything, with minimal automation.
- Level 1 (Driver Assistance): Offers basic automation features like adaptive cruise control or lane-keeping assistance, but the driver must remain in control.
- Level 2 (Partial Automation): Vehicles can control both steering and acceleration/deceleration in certain conditions, but the driver must still be attentive and prepared to take over.
- Level 3 (Conditional Automation): The vehicle can manage most driving tasks under specific conditions, but the driver needs to be ready to intervene if necessary.
- Level 4 (High Automation): The vehicle can operate autonomously in certain environments and conditions, but human intervention may still be needed in some situations.
- Level 5 (Full Automation): The vehicle can manage all driving tasks in all conditions without any human intervention, which is currently still in development.
While Level 5 AVs are currently available for general sale, many cars on the road today offer Level 2 partial automation, with some manufacturers also offering or testing Level 3 and even Level 4 capabilities in specific situations.
The Long, Winding Road To Level 5 Autonomy
Level 5 automation could be decades away due to technological hurdles, the need for extensive testing, and the validation of complex scenarios to ensure safety and reliability. For example, developing sensors and algorithms that can accurately interpret complex driving environments under various conditions (weather, lighting, traffic) is a significant challenge. So, too, is creating reliable software that can make safe decisions in unpredictable situations and learn from new experiences.
Ensuring effective communication between vehicles and infrastructure poses a significant obstacle, joined by the power needed to process the sheer amount of real-time data an AV needs to collect and analyze in split seconds.
Then there are the challenges of testing and validating the decision-making software algorithms across billions of driving scenarios and designing systems with redundancy and fail-safe to ensure that the vehicle can manage unexpected situations or malfunctions.
Those are obstacles related to designing Level 5 AVs, but there are human challenges to overcome, too. These include building public trust and addressing ethical dilemmas related to autonomous driving, such as who is responsible when an accident occurs.
Consideration of infrastructure has to be made to ensure roads are compatible with autonomous vehicles, such as having clear markings and traffic signals. And, finally, developing appropriate regulations and legal frameworks for AVs remains a work in progress.
Despite the challenges, PatentPC writes, “More than 40 companies are investing billions of dollars into AV technology, hoping to be the first to achieve Level 5 autonomy. Tesla, Waymo, Cruise, Baidu Apollo, and Mercedes-Benz are among the top players, each taking different approaches.”
The reason so many companies are spending so much money is clear: according to PatentPC, “The self-driving industry is expected to become a trillion-dollar market in the coming decades, driven by advancements in technology, regulatory support, and consumer demand for mobility solutions.”
With so much money at stake, the need to overcome the above challenges remains paramount. So, too, does the need to address one other concern: what to do with the voluminous amount of data being collected by AVs?
* For more information about the levels of driving automation check out this SEA graphic.
What To Do With All That Data
The road to Level 5 automation is influenced by concerns surrounding data collection. These include privacy, security, regulatory compliance, and ethical considerations, all of which are pivotal in shaping the deployment and public acceptance of AVs.
AVs rely on an array of sensors and cameras to navigate, resulting in the continuous collection of vast amounts of data. This data includes not only vehicle performance metrics but also sensitive information such as precise location tracking, video recordings of surroundings, and potentially personal details of passengers and pedestrians.
The Electronic Frontier Foundation has highlighted concerns regarding the potential for this data to be exploited, emphasizing the need for robust privacy protections to prevent unauthorized surveillance and misuse.
Moreover, the integration of connected vehicle technologies introduces vulnerabilities that could be exploited by malicious actors. The Financial Times reported that as vehicles become more connected, they pose real risks related to cybersecurity and data privacy, necessitating stringent measures to safeguard against potential breaches.
In response to these concerns, regulatory bodies are taking action to mitigate potential threats. The Biden administration, for instance, proposed rules to ban the sale or import of connected vehicles equipped with technology from countries deemed as security risks, such as China and Russia. This initiative aims to prevent potential espionage and protect national security by restricting components that could be exploited for unauthorized data collection or remote manipulation.
These regulatory measures underscore the complexities automakers face in ensuring compliance while fostering innovation. A survey conducted by KPMG revealed that 86% of automotive leaders have significantly increased their privacy program budgets, reflecting the industry’s recognition of the importance of data privacy in building consumer trust.
The ethical implications of data collection by AVs extend to issues of surveillance and individual freedoms. The potential for AVs to be used as tools for mass surveillance raises significant ethical questions. The Electronic Frontier Foundation warns that the vast amount of data collected by self-driving cars could be accessed by law enforcement or other entities, potentially infringing on civil liberties.
Liability concerns are also paramount. In the event of an accident involving an AV, determining fault becomes complex, especially when considering the role of data in reconstructing events. The retention and accessibility of this data are critical for legal proceedings, yet they also pose risks if mishandled or accessed without proper authorization.
Impact On Deployment And Public Trust
These multifaceted data concerns contribute to delays in the widespread deployment of all levels of AVs. Automakers must navigate a landscape where technological innovation intersects with stringent data protection requirements and public skepticism. Building and maintaining consumer trust is contingent upon demonstrating a commitment to privacy and security.
The KPMG survey indicates that while the automotive industry is still in the early stages of adopting leading privacy practices, significant strides are being made to address these challenges.
Data collection concerns are integral to the discourse on AVs, influencing regulatory frameworks, ethical debates, and the pace of technological adoption. Addressing these issues is essential for the responsible and accepted integration of self-driving cars into society.