The idea of a car that drives itself once belonged to science fiction. Today, it is moving closer to everyday reality, with pilot programmes and trials already running in cities around the world. Yet alongside the excitement come significant challenges, from regulation and insurance to public trust and cost. To understand where the industry is heading, it helps to begin with the basics.
A self-driving car, also known as an autonomous vehicle or driverless car, is a vehicle designed to operate with little or no human intervention. Using a combination of sensors such as cameras, radar, and lidar, together with artificial intelligence and real-time data processing, these vehicles can perceive their environment, interpret traffic conditions, and make safe driving decisions.
Depending on its level of automation, a self-driving car may still need a human driver in certain situations, or it may be capable of fully autonomous travel without the use of a steering wheel or pedals.
SAE levels of driving automation
To bring some structure to the conversation, the Society of Automotive Engineers created a the SAE 5 levels of driving automation scale.
SAE Level | Description | Example Use Case |
|---|---|---|
0 | No automation, driver does everything | Manual Driving |
1 | Driver assistance (e.g. cruise control) | Speed control on motorway |
2 | Partial automation; steering and acceleration handled, but driver must stay alert | Tesla Autopilot, Ford BlueCruise |
3 | Conditional automation; car drives itself in certain conditions but may ask for takeover | Honda Sensing Elite in Japan |
4 | High automation in geofenced areas | Waymo robotaxis in Phoenix |
5 | Full automation everywhere, no human controls needed | Still conceptual |
Autonomous Vs automated Vs self-driving
Although often used interchangeably, these terms carry slightly different meanings. “Autonomous” suggests total independence, implying that a vehicle can make decisions without human instruction. “Automated” is the precise term preferred by engineers, focusing on replacing driving functions without implying intelligence. “Self-driving” is the most common public phrase and usually refers to cars that can operate independently in some conditions but still require a human driver in others.
How do self-driving cars work
At the heart of every autonomous vehicle lies a fusion of advanced hardware and intelligent software. Cameras and radar measure the movement and distance of surrounding vehicles, lidar builds a three-dimensional map of the environment, and ultrasonic sensors help detect nearby objects and assist with parking.
These streams of data are then processed by high-performance onboard computers. Artificial intelligence and machine learning models integrate the inputs to create a live map of the surroundings, classify objects, and predict the movements of pedestrians, cyclists and other vehicles. Based on this analysis, the system makes continuous decisions and sends commands to the car’s controls to steer, accelerate and brake. In effect, the car is constantly perceiving, deciding and acting in fractions of a second.
Benefits of self-driving cars
- Improved safety: Human error causes most crashes. Automated systems do not get tired, distracted or impaired, and early trials (such as Waymo’s) show lower accident rates than human drivers.
- Accessibility and mobility: Elderly people and those with disabilities could gain new independence. Robotaxis and automated shuttles may provide affordable transport in areas underserved by public transit.
- Efficiency gains: Fleets of autonomous vehicles can operate for longer hours without rest, reducing labour costs and improving logistics and delivery schedules.
- Environmental benefits: Combined with electrification and shared use, AVs can reduce emissions, ease congestion and reshape urban planning by lowering the demand for parking.
Challenges of self-driving cars
- Technological limitations: Current systems still struggle in poor weather, on unusual road layouts and in complex urban traffic. Sensors are costly, and AI has difficulty with unpredictable “edge cases.”
- Legal and ethical issues: Responsibility in the event of a collision remains unclear. Ethical dilemmas, such as unavoidable harm scenarios, are unresolved.
- Public trust and perception: High-profile accidents have raised doubts. Surveys show many people remain sceptical, even when safety data is positive.
- High costs: Development and repair of AVs is expensive, meaning early adoption will focus on fleets and commercial services rather than private ownership.
Companies driving the autonomous vehicle industry
Key players at a glance
- Waymo (Alphabet) – Leading robotaxi operator in US cities
- Cruise (GM) – Robotaxi developer, currently paused after safety incidents
- Tesla – Level 2 “Full Self-Driving” system for consumers
- General Motors (Super Cruise) – Advanced driver assistance system
- Ford (BlueCruise) – Consumer-level driver assistance
- Gatik – Specialising in middle-mile autonomous freight
- Kodiak Robotics – Long-haul trucking automation
- Einride – Cabless, remote-operated electric freight pods
- Zego – Providing car insurance for self-driving cars, supporting AV trials and future fleets
Consumer systems
On the consumer side, the most advanced systems available today remain at Level 2. General Motors’ Super Cruise, Ford’s BlueCruise and Tesla’s Full Self-Driving all allow cars to steer, accelerate and brake under specific conditions, though drivers must remain fully engaged. These systems represent a stepping stone rather than true autonomy, but their rapid adoption shows the appetite for automation among car buyers.
Robotaxis
At the frontier are companies developing vehicles that go far beyond driver assistance. Waymo, a subsidiary of Alphabet, continues to lead with robotaxi services already running in Phoenix, San Francisco, Los Angeles and Austin. Its safety record has been cited as evidence self driving cars can be safer than humans under the right conditions. Cruise, owned by General Motors, has faced setbacks after suspending its robotaxi operations following safety incidents, but it remains one of the highest-profile players in the space.
Freight and logistics
In freight and logistics, where routes are predictable and economic incentives are strong, progress is accelerating. Gatik specialises in middle-mile trucking, transporting goods between distribution centres and retail locations. Kodiak Robotics is focusing on long-haul and local freight with trucks designed to operate driverless on motorways. Swedish company Einride has taken a radical approach with cabless, remote-operated electric pods that already deliver goods across Europe and North America.
Insurance and risk management
Alongside the manufacturers, insurers play a vital role in making autonomous vehicles possible. Without clear cover and liability frameworks, trials cannot move forward safely. Zego is one of the first companies promising to offer car insurance for self-driving cars, supporting early trials and preparing for future fleets. By using driving data and telematics, Zego helps prove that autonomous vehicles meet safety standards, giving operators, regulators and the public the confidence to trust this new technology.
The Road Ahead: What’s Next for Self-Driving Cars?
The future of self-driving cars is not arriving overnight, but the momentum is undeniable. According to McKinsey & Company, autonomous driving could generate between $300 billion and $400 billion in annual revenue by 2035, largely from shared mobility services, logistics fleets and premium driver assistance features. They also suggest that robotaxis and shared AV transport could eventually replace a significant share of private car journeys, easing congestion and altering the way cities are designed.
Goldman Sachs adds that by 2030 around 10% of new cars sold could have Level 3 automation, where the vehicle can handle most driving tasks under defined conditions, while roughly 2.5% may reach Level 4, offering full autonomy in specific areas. At first glance these numbers may look small, but when you consider the scale of the global car market, it still represents millions of self-driving vehicles on the road within a decade.
From cost to confidence
The economics of autonomy are also shifting quickly. LiDAR sensors once costing tens of thousands of dollars each, are forecast by several analysts to drop below $200 at scale, bringing the cost of AV technology down to levels comparable with other mainstream automotive features. This drop, combined with progress in AI and computing power, makes widespread deployment far more realistic, particularly for commercial fleets where uptime and efficiency are critical.
Yet technology and cost are only part of the story. Trust, regulation and insurance will shape how quickly adoption happens. The UK’s Automated Vehicles Act 2024 requires autonomous vehicles to meet the safety standard of a “competent and careful driver.” Meeting that benchmark will be vital not just for legal approval, but for convincing the public that driverless technology is safe. This is where companies like Zego play a role: by offering car insurance for self-driving cars, supporting early trials and future fleets, and using telematics to measure and prove safety standards. Without these frameworks, the industry cannot scale.
Looking further ahead, the impact on everyday life could be profound. Fewer road accidents, reduced congestion, and cleaner air are all realistic outcomes. Cities may reclaim space from parking and traffic, dedicating it instead to people, parks and housing. Logistics could become faster and more reliable, with deliveries running 24/7. And for many people, private car ownership may no longer be essential, replaced by on-demand autonomous transport services that feel as natural as booking a ride-hailing car today.
Frequently Asked Questions About Self-Driving Cars
Self-driving cars spark a lot of curiosity, and confusion. From safety and insurance to timelines and technology, people want clear answers. Below, we’ve covered the most common questions with data-backed insights to help you understand where autonomous vehicles stand today and what’s coming next.
Who is responsible if a self-driving car crashes?
When a self-driving car is involved in an accident, the question of liability becomes complex. Responsibility could fall to the human safety driver (if one is present), the car manufacturer, or even the software provider. The UK’s Automated Vehicles Act 2024 says autonomous cars must drive to the standard of a “competent and careful driver,” but it doesn’t yet spell out exactly how blame and costs will be shared. Until global rules catch up, insurance is the bridge that makes trials and early rollouts possible.
(gov.uk, techdigest.tv)
Can self-driving cars really reduce accidents?
Human error is responsible for the majority of crashes, so automation has huge potential to make roads safer. Research from IDTechEx shows that 99% of autonomous vehicle crashes involved human behaviour, not the AV systems themselves. At the same time, the Insurance Institute for Highway Safety estimates that automation might only prevent around one-third of all collisions, since not every scenario is avoidable. The evidence suggests AVs could lower accident rates overall, but won’t eliminate crashes completely.
(iotworldtoday.com, axios.com)
When will self-driving cars be available in the UK?
The UK government originally expected self-driving cars to be on roads by 2026, but rollout has been delayed. Current guidance suggests 2027 is a more realistic date for wider adoption, beginning with trials and commercial fleets before private ownership. This slower timeline reflects the need for more safety testing, clearer liability frameworks, and public trust.
(techdigest.tv, weforum.org)
Are self-driving cars safer than human drivers?
Studies suggest they could be safer, but the answer isn’t simple. Around 94% of serious crashes are caused by human mistakes, from distraction to fatigue. Self-driving systems don’t get tired or drunk, but they can struggle in bad weather or unusual road layouts. Research indicates automation could only prevent about 34% of crashes, meaning AVs will reduce risk but not remove it.
(lookupaplate.com, axios.com)
How many accidents involve self-driving cars today?
In the U.S., the NHTSA has logged nearly 4,000 autonomous vehicle-related incidents since 2021, resulting in 83 deaths, though most were minor collisions. California alone recorded 100+ AV crashes since 2023, largely non-fatal bumps during testing. These figures show that AVs are not immune to accidents, but the majority have been low-severity compared to traditional car crashes.
(craftlawfirm.com, gmdirecthire.co.uk)
Do self-driving cars need insurance?
Yes. Just like conventional vehicles, autonomous cars must have self driving car insurance before they can legally operate. The difference is that car insurance for self-driving cars also has to cover complex liability questions between drivers, manufacturers and software providers. That’s why insurers like Zego are developing dedicated products to support early trials, AV fleets, and compliance with the UK’s Automated Vehicles Act 2024.
