Overview of Could we have robotaxis by next year?
ABC News Daily explores how close autonomous taxis are to becoming a reality in Australia, using London’s new robotaxi trial as the springboard. In conversation with robotics expert Professor Michael Milford from QUT, the episode explains how the technology works, why London is a major test case, and what still needs to happen before driverless taxis can operate widely in Sydney, Melbourne, or other Australian cities.
What’s happening in London
London is now seeing a small-scale robotaxi trial, with a vehicle operating under human supervision for safety. The episode notes that the city is a tough proving ground because of:
- narrow, ancient streets
- heavy traffic
- cyclists, scooters, pedestrians, buses, and roadworks
- unpredictable urban driving conditions
The London trial is framed as a major step, but still very early-stage.
Why London matters
The guest explains that if a robotaxi can handle central London, it can likely handle many other complex cities. London and San Francisco are seen as some of the hardest driving environments for autonomous vehicles.
How robotaxis work
Milford says the technology is highly advanced and has been in development for decades. It relies on two major elements:
- sensors that let the car perceive its surroundings
- AI systems that make driving decisions in real time
A key point is that some companies rely less on detailed pre-mapped roads and more on the vehicle’s ability to learn to drive from data and experience. This makes their approach more like a human learning process, though still far from perfect.
Ride quality and human behavior
He also notes that developers have spent years improving ride smoothness, because passengers are sensitive to unnatural acceleration and braking. In that sense, robotaxis can sometimes feel smoother than a human driver.
What the biggest challenges are
The episode spends a lot of time on the remaining obstacles to widespread deployment.
Safety and edge cases
Autonomous vehicles still struggle most with unexpected events, such as:
- a van suddenly opening its door
- pedestrians appearing unexpectedly
- unusual road behavior that may require the car to “bend” driving rules slightly, the way humans sometimes do
The transcript also references high-profile incidents overseas, including collisions and animal deaths, as reminders that the technology is not flawless.
Regulation and standards
A major issue is that safety data from the big autonomous-vehicle companies is:
- published in different formats
- difficult to compare
- not standardized enough for regulators to easily judge overall safety
Milford argues that regulators need clearer evidence that robotaxis are safer than human drivers, not just “different.”
Liability
If there is no human driver actively controlling the vehicle, responsibility becomes complicated. Liability could fall on:
- the operator
- the manufacturer
- or be sorted out through courts over time
What this means for Australia
The episode says Australia is getting closer, but not immediately.
Possible timeline
Milford suggests:
- small-scale Australian trials could happen in the near future
- legal deployment may become possible around 2027
- widespread adoption will take longer
He says there is no technical reason robotaxis couldn’t work in Australia, as long as companies invest enough time and data to adapt them to local conditions.
Australian-specific hurdles
A few local concerns are mentioned, including:
- left-side driving
- hook turns
- kangaroos and other unpredictable wildlife
Milford’s view is that these are not unsolvable problems — they mostly require more local training, testing, and investment.
Broader impact: jobs, acceptance, and realism
The episode also touches on the social side of robotaxis.
Job disruption
Taxi drivers and the wider transport industry are likely to resist change because of job impacts. Milford says this needs to be managed with:
- retraining
- upskilling
- transition planning for affected workers
He also notes that the taxi industry was already heavily disrupted by Uber.
Not inevitable, but likely
His overall position is balanced:
- robotaxis are not inevitable
- society can slow or shape adoption
- but the potential to reduce deaths caused by human driving error is a major reason the technology is still compelling
Bottom line
Robotaxis are no longer science fiction, and Australia could see early trials sooner than many people expect. But the episode makes clear that full-scale rollout is still limited by regulation, safety proof, liability questions, and public trust. Milford’s view is cautiously optimistic: small trials in Australia soon are plausible, but a future where most taxi rides are driverless is likely further away than one decade.