Ever since Tesla announced its self-driving features in 2014, the company has been a leader in the progression of autonomous vehicles.
Their self-driving technology combines software that controls automated steering, acceleration, and braking on roads and highways with lanes. The strides Tesla has made with autonomous driving technology do not come without stepbacks. Since Tesla’s autopilot inception, there have been multiple accidents, including three fatal ones.
While Tesla fully backs their autopilot feature, experts say the data used to back up these claims is incomplete. To learn more about Tesla’s autopilot failure of the decade, keep reading.
Tesla’s Response to Autopilot Fails
When authorities in Texas concluded that ‘no one was driving’ in the Model S that crashed into a tree in April 2021, CEO Elon Musk tweeted “autopilot was not enabled” with a series of logs.
Another crash in Florida that resulted in fatalities included a Tesla engaged in autopilot. Tesla stated autopilot is “designed for use on limited-access highways with no cross traffic.” They also reiterate that the manual clearly says the autopilot feature is a driver-assist system, where the driver must have their hands ready on the wheel to be able to take control of the car.
Tesla’s responses to deadly accidents involving the autopilot feature leave consumers with questions and concerns.
Tesla’s Perception of Autopilot Failures
Tesla in the media portrays and describes the autopilot feature as your idea of the future: a self-driving car. However, this is not the case.
Tesla autopilot is a driver-assist software, meaning it does just that: it assists the driver. On Tesla’s website, it states, “autopilot and Full Self-Driving Capability are intended for use with a fully attentive driver, who has their hands on the wheel and is prepared to take over at any moment. While these features are designed to become more capable over time, the currently enabled features do not make the vehicle autonomous.”
They also express that autopilot can malfunction in unacceptable conditions, such as poor visibility, including weather like rain, snow, fog, overwhelmingly bright lights from oncoming traffic, or direct sunlight. Autopilot can also be affected by obstructions caused by paint or road repairs and narrow, high, winding roads.
The hardware that helps autopilot operate includes a series of high-tech cameras and sensors that send alerts about the environment and any potential obstructions around the car. Tesla’s support website says these cameras can experience dirt buildup from car use and need to be cleaned with a damp cloth.
Tesla is no stranger to criticism from consumers and the media. The rise in the popularity of Tesla vehicles is due to its all-electric appeal and notoriety in the self-driving market.
With fame comes criticism, especially after multiple crashes and fatalities while autopilot had been engaged.
Musk has feverishly denied any error in the Tesla autopilot software, blaming irresponsible drivers for any incidents behind the wheel.
In the fatal crash in Houston, authorities only found victims in the passenger seat and backseat, but no one was driving the Tesla. Musk stated the data logs from Tesla show autopilot was not engaged, but authorities say there’s no one who could have switched their seat after the crash occurred.
Tesla is one of the only car manufacturers that have autonomous driving available for consumer use. If other competitors made their autonomous diving software available on new vehicles, it begs whether there would be more or less criticism.
What Caused Tesla Autopilot Failures?
The fatalities in the Tesla autopilot crashes are alarming and begs the question: What caused them?
For the Texas and Florida accidents, authorities aren’t quite sure. Tesla is adamant autopilot is not to blame. So what went wrong? In many crash reports, Tesla blames the driver of the vehicle. In a 2020 crash in California, Tesla blamed an autopilot failure on account of the driver, Walter Huang, stating, “the only way for this accident to have occurred is if Mr. Huang was not paying attention to the road.”
Even though Tesla’s cars are not fully self-driving, comments made by the company, advertising efforts, and media promotions cloud this fact. Many statements by the company, including Tesla’s CEO Elon Musk, would lead a consumer to believe otherwise.
The technical capabilities of the autopilot are limited as well. When the autopilot is engaged, outside factors can affect the cameras and sensors to detect obstructions, stopped vehicles, and pedestrians. These factors can include bright sunlight or uneven or no lane markings.
It’s hard to say what caused the failures, but authorities believe a malfunction by the car and delayed react time from the driver could be possible. When consumers are relying so heavily on autopilot, mistakes are bound to happen.
The Race to Safer Autonomous Driving
When it comes to autonomous vehicles, Tesla has a long list of competitors. And because Tesla is the face of electric vehicles in so many aspects, and their failures are so heavily publicized, these competitors are able to learn from Tesla’s mistakes and make strides towards safer autonomous-driven vehicles.
By the year 2040, it has been estimated there will be over 30+ million driverless vehicles on roads. The market for self-driving vehicles is sitting at $54 billion.
These top twelve companies have made strides in autonomous driving and are leading the market.
Google’s own self-driving car company, Waymo, was created in 2009. In 2015, Waymo made a car without a steering wheel, brake, or gas pedal and achieved the world’s first fully self-driving car on any public road in Nevada.
In 2018, Jaguar Land Rover supplied Waymo with 20,000 cars to be converted to fully self-driving vehicles for its ride-hailing service. In 2019, Renault-Nissan-Mitsubishi announced the same, to partner with Waymo to help create a fully autonomous ride-hailing system.
In 2020, Waymo announced their autonomous vehicles had driven ten million miles on public roads in a year. They also partnered with UPS to deliver packages in autonomous trucks across Arizona.
The progress they’ve made in the last ten years has increased quickly in the previous five, showing the incredible growth of the autonomous vehicle market.
Founded in 2016 by Allstate, Arity uses data to identify traffic patterns to predict and reduce accident risk. The data collected can also advise drivers about insurance and give automotive companies insight into how to minimize the risk in their self-driving capabilities.
To date, Arity has analyzed about 30 billion driving miles. The data collected examines how humans interact with other humans while driving, how often we look at our phones, how humans interact with self-driving cars, and even how vehicles interact with each other.
Arity says all of this available data will only benefit the future of transportation. It gives valuable insight that’s used to develop better and safer practices for autonomous vehicles.
What do you need to power these self-driving electric vehicles? Chargers and power stations.
Founded in 2007, WiTricity achieves just that. WiTricity develops wireless charging stations for electric vehicles, giving drivers ease of mind to know they’ll always have a power source.
WiTricity is genuinely paving the way for wireless car charging. WiTricity technology allows you to pull into a WiTricity parking spot at a charging station and charge your Electric Vehicle without plugging anything in.
With fully autonomous cars, this diminishes the need for any human intervention to plug in a power cord at the end of its journey. It just needs to find its way to a WiTricity parking station.
Unity Technologies, founded in 2004, helps develop 3D and Virtual Reality (VR) visualization tools to help companies develop autonomous driving capabilities.
Unity’s engineering teams offer simulation environments for autonomous vehicle training that adequately replicate experiences that include sensory and physical complexity, cognitive challenges, and support intelligent interactions.
Unity provides excellent tools and enrichment for the autonomous vehicle training that will provide information to make these self-driving vehicles more intelligent and versed in the way humans operate and interact with other cars and environments.
Ouster was founded in 2015 and created 3D lidar sensors for use in autonomous vehicles. These sensors have technology that gives autonomous cars a better sense of their surroundings.
Ouster creates these Lidar sensor systems that give autonomous cars a virtual map of their surroundings. This system provides a vehicle with an accurate picture of what’s around the vehicle to respond accordingly. Lidar sensors have a high level of accuracy and reliability.
The engineers developing this technology are furthering advancements in autonomous driving capabilities every day. They’ve made impressive strides in mapping technology and car security systems, all of which can be used to promote the development of autonomous vehicles.
Cruise, founded in 2013, aims to help connect riders with autonomous vehicles for ride-sharing on demand. They are a leader in cutting-edge robotics and AI technology.
Cruise has also partnered with Walmart to deliver groceries with their autonomous vehicles in Phoenix, Arizona.
Their vehicles use a series of sensors to help the car see what’s around them. The technology developed by Cruise helps their autonomous cars predict and react to animals, people, and objects on or around the road. Cruise’s vehicles have logged around two million miles on the streets of San Francisco alone.
General Motors (GM) announced in 2018 they would start production on self-driving electric vehicles. They began production on self-driving electric Chevy Bolts, which GM calls the “Cruise AV.”
In 2019, GM, Ford, and Toyota joined SAE International to compose safety standards for autonomous cars, which would eventually help regulate all autonomous vehicles in the United States.
GM is one of the first major car companies to step towards electric vehicles and implement that technology to their current production. GM owns Cruise, its autonomous vehicle subsidiary, which also acquired Voyage. Together, they developed an electric self-driving vehicle called Origin, which lacks any steering, gas, or brake pedals.
GM CEO Mary Barra sees GM selling personal autonomous vehicles to consumers by the end of 2030.
GM launched Super Cruise in 2017, which is a semi-autonomous driver assist feature on specific Cadillac models. More advanced versions of Super Cruise are set to roll out around 2023.
Amazon’s contribution to the autonomous vehicle market is Zoox. This startup was created in 2014 and focuses on developing vehicles that can drive on city streets and highways while maintaining the safety of passengers.
Zoox’s fully electric autonomous cabs include a bidirectional driving capability, which reduces the need to reverse and consists of four seats with fully equipped airbag systems. With 1063 employees, Zoox is currently the fourth largest company working on autonomous vehicles in San Francisco.
Like other autonomous car companies, Zoox aims to create a fleet of fully autonomous cars that can independently operate in cities as transportation options.
Toyota (Woven Planet)
Toyota is another classic automotive company that has a stake in the autonomous driving market. As of 2020, Toyota spends roughly ten billion a year on autonomous vehicle research.
Toyota has worked with leaders in mapping technology to help develop software that will make autonomous vehicles read and perform better in high-traffic areas. In March 2020, Toyota, TomTom & Denso worked together to build high-definition maps to improve and implement groundbreaking technology in autonomous vehicles.
In April of 2021, Toyota’s subsidiary, Woven Planet, bought Lyft’s autonomous driving subsidiary. This acquisition will significantly benefit the development of fully autonomous vehicles for ride-sharing services.
Levels of Autonomous Driving
With the development of autonomous vehicles, a ranking system was implemented by the Society of Automotive Engineers (SAE) that ranks levels of autonomous cars from level 0 (fully manual cars) to level 5 (fully autonomous).
Level 0 cars have no driving automation. Level 0 vehicles are fully manual and controlled by the driver.
Level 1 autonomous vehicles include any car with a driving assistance feature, like cruise control. Drivers are still in control of the car and can manually use steering, braking, and gas pedals.
Level 2 is partial driving automation. Tesla’s vehicles sit at level 2. At this level, the vehicle can control braking and acceleration, as well as steering mechanisms. A driver who sits in the driver’s seat of the car can take control at any time.
Level 3 autonomous vehicles make decisions based on their surrounding environment. These level 3 ranked vehicles have a system of sensors and cameras to help assist and react. These level 3 vehicles still require a driver to be at the ready to take over if the system fails.
Level 4 autonomous vehicles are similar to level 3, except the vehicle can intervene if the system fails. Many autonomous ride-sharing companies discussed earlier have developed testing vehicles with level 4 autonomy.
This level is full driving automation. These vehicles do not require any human to be in the driver’s seat. Level 5 autonomous vehicles will be manufactured and function without a steering wheel or brake and gas pedals.
As of 2021, no level 5 fully autonomous vehicles are available to the general public; however, many are undergoing testing. Tesla’s vehicles are currently ranked at Level 2 Autonomous.
The United States still has a ways to go when it comes to level 5 fully autonomous vehicles. The technology and production aspects are promising, but the lack of regulations and security keeps these vehicles from becoming mainstream.
Common Challenges of Autonomous Vehicles
As the market for autonomous cars continues to develop, many issues seem to arise. The technology for self-driving vehicles is advanced but also ever-changing.
Some common problems among autonomous vehicles include creating and maintaining maps. Creating and maintaining 3D maps of suburban areas for autonomous cars involved in ride-sharing services can help them predict and react in busy intersections and crowded places.
Relying on this map system limits the number of cities autonomous cars can operate, and updating them is time-consuming.
Technology for autonomous vehicles still can’t accurately react in certain situations. Many aspects of driving are social interactions, which technology is still not very good at gaging, like instances with pedestrians and cyclists.
The Future of Autonomous Vehicles and Tesla
Tesla’s autopilot failures remind us of another essential part of the autonomous vehicle market. Regulations and security are a more extensive conversation that needs to be discussed before approving fully autonomous vehicles to be available to the public.
Cybersecurity is a concern as well. There needs to be a system in place to prevent autonomous vehicles from being hacked. As of 2021, there are no specific regulations about autonomous vehicles and cybersecurity. The United States Department of Transportation has created legislation based on autonomous vehicles, mostly leaving those decisions to the states.
As the technology develops further, we will see more laws about autonomous vehicles, especially as more companies expand into the autonomous vehicle market. Creating regulations will eventually help diminish the potential for hacking and injury.
As Tesla develops and betters its autopilot feature to keep up with the market, inevitably, these incidents and failures will not be the last. As we see an increase in autonomous driving, issues are bound to happen.