Electric Vehicles
The
electrifying technology behind Electric Vehicles (EVs), specifically Tesla
models, have significantly driven the transition towards solutions for
automated transportation (Armstrong, 2023). The Autopilot system, developed by
Tesla, is considered by many to be the forefront of vehicle automation, as it
leads the charge for the future of cars and transportation by merging
cutting-edge hardware with advanced software. Essentially, Autopilot is an
advanced driver assistance system (ADAS) that enhances safety and convenience
when driving.
At
its core, Autopilot employs 8 external cameras, ultrasonic sensors, radar, and
a powerful onboard computer to perceive and interpret the surrounding
environment (Tesla, n.d). The key features of Autopilot include Traffic-Aware
Cruise Control (TACC) that adjusts the cruising speed of the vehicle to match
surrounding traffic, Autosteer to assist in steering on vehicular roads, and
Navigate on Autopilot, which enables automated lane changes and highway
interchanges. In addition to these driving-assist features, Autopilot also
includes several advanced safety technologies, such as Automatic Emergency
Braking, Forward and Side Collision Warning and Obstacle Aware Acceleration
(Tesla, n.d). Although a driver must remain behind the wheel, Autopilot allows
the vehicle to drive autonomously with minimal driver intervention, reducing
the overall workload of a driver through partial vehicle automation (Tesla,
n.d). Despite being a revolution in the car industry, Tesla's Autopilot is plagued by persisting
limitations in certain essential areas, such as not being able to be fully
autonomous, having navigation capabilities impacted by poor visibility and
inaccurate and outdated map data.
Tesla Autopilot’s most consequential limitation is its
unavoidable need for a driver to consciously operate the vehicle. As opposed to
manual driving, Autopilot gives drivers the illusion that the vehicle is
entirely self-driving, which inadvertently encourages distracted driving –
leading to accidents and fatalities. Furthermore, Autopilot sometimes fails to
recognise other vehicles and stationary objects, resulting in accidents as the
vehicle is unable to stop when required. According to Boudette’s (2021) article
which detailed a fatal Tesla Autopilot accident, a driver’s lapse in attention
resulted in a fatal crash with the car in front. As such, it is paramount to
have safeguards in place to ensure that drivers stay attentive, should they be
required to reclaim manual control as situations arise (Casner et al., 2016).
Poor visibility caused by weather conditions and
dirty cameras will also affect Autopilot’s operational capabilities. As
Autopilot employs 8 external cameras to perceive the surrounding environment,
sub-par weather conditions like heavy rain or snow, that may result in cameras
and sensors being covered or dirtied, would heavily limit the navigation
capabilities or even render the navigation ability useless. According to
Iqtidar Ali’s (2019) article on Tesla’s Smart Summon function during light
snowfall, Autopilot was only able to navigate to its owner under snowy weather
conditions at a speed of 1 mph because its Smart Summon detected the weather
conditions to be unideal. In more severe conditions that fully cover the
external cameras, the Smart Summon feature cannot be used and the vehicle would
remain stationary.
Another limitation of Autopilot is that it is heavily
dependent on updated data in its Global Positioning System (GPS) to navigate
accurately and safely. Inaccurate and outdated map data will result in
incidents where Autopilot drives through new stop lights and stop signs not
reflected in its GPS. According to Jotti Mann (2023), a Tesla Model 3 in Full Self-Driving mode “did not slow down
before it hit water on the road and crashed into a pond”. The road had a
temporary "flooded" sign on the side of the road alerting drivers to
the water ahead, which Autopilot failed to capture and account for, as it was
not reflected in its GPS.
Though Autopilot has limitations which could cause
fatalities if not addressed properly, one must consider that there is
prospective use in reducing the workload of a driver because less focus is
needed to control the vehicle with Autopilot’s automation features like TACC,
Autosteer, and Navigate on Autopilot. An experiment done by Transportation Research Group in
the University of Southampton showed that Autopilot could reduce the workload
of a driver to overall low levels of between 10% to 43% (variable environment
being relatively empty highways to busy city rings)
In conclusion, Autopilot should be treated as an
advanced driving aid rather than an autonomous driving system. The
functionalities of Autopilot serve to enhance safe driving and assist drivers but
should not instill an unwarranted sense of confidence. In short, even with
Autopilot, drivers still need to remain alert to their surroundings and be
prepared to take over the wheel at any time. That said, continuous advancements
in technology, such as better map data and sensor reliability, will also mitigate
its limitations and weaknesses. Ultimately, the objective is to leverage the
benefits of automation while ensuring safety on the road.
References
Ali, I. (2019). Watch Tesla Model 3
performing Smart Summon during light snowfall. From
https://www.xautoworld.com/tesla/model3-smart-summon-snowing/.
Armstrong, K. (2023). Tesla Pioneers the
Future: How Autonomous Cars Will Transform Transportation.
https://www.notateslaapp.com/tesla-reference/1353/tesla-pioneers-the-future-how-electric-and-autonomous-cars-will-transform-car-ownership-and-transportation.
Boudette, N. E. (2021). ‘It Happened So
Fast’: Inside a Fatal Tesla Autopilot Accident.
https://www.nytimes.com/2021/08/17/business/tesla-autopilot-accident.html.
Heikoop, D. D., Winter, J. C., Arem, B. v.,
& Stanton, N. A. (2019). Acclimatizing to automation: Driver workload and
stress during partially automated car following in real traffic.
https://www.sciencedirect.com/science/article/abs/pii/S136984781830500X.
Mann, J. (2023). A Tesla owner said his
Model 3 in Full Self-Driving mode didn't slow down before it hit water on the
road and crashed into a pond.
https://www.businessinsider.in/thelife/news/a-tesla-owner-said-his-model-3-in-full-self-driving-mode-didnt-slow-down-before-it-hit-water-on-the-road-and-crashed-into-a-pond/articleshow/103106256.cms.
Stephen M. Casner, Edwin L. Hutchins, & Don Norman (2016).
The Challenges of Partially Automated Driving. Communications of the ACM.
https://cacm.acm.org/magazines/2016/5/201592-the-challenges-of-partially-automated-driving/fulltext.
Tesla. (n.d). Autopilot. https://www.tesla.com/support/autopilot.