Otomedik

Otomedik · Guide

Autonomous Robotaxi Services: A Guide to Safety, Oversight and Passengers

As driverless robotaxi services expand, passengers, fleet operators and vehicle buyers need to assess technology claims, safety oversight and responsibilities correctly. This guide explains how to approach autonomous vehicles without steering wheels or pedals in practice.

September 05, 2026Otomedik
Share
Autonomous Robotaxi Services: A Guide to Safety, Oversight and Passengers
Contents

The idea of a driverless robotaxi is one of the automotive industry's most discussed transformations. Calling a vehicle through an app, getting in, and reaching a destination without a human driver creates important possibilities for accessibility, urban mobility and fleet efficiency. Yet putting a vehicle with no steering wheel, brake pedal or conventional driver position on the road is not merely a software achievement; it is also a matter of vehicle safety, operational discipline, regulation, maintenance and emergency management.

Announcements surrounding concepts or limited service fleets without steering wheels and pedals, such as the Tesla Cybercab, make that discussion more visible. A company's launch of a service in a particular city, or an administrative review concerning vehicle certification, is not in itself proof that a vehicle is either safe or unsafe. For passengers, investors, rental companies and used-car buyers, the right approach is to look beyond headlines: examine the conditions in which the service operates, who monitors it, and what happens when something goes wrong.

The difference between a robotaxi and advanced driver assistance

The most important distinction is between a driver-assistance system and a driverless transportation service. With lane centring, adaptive cruise control or advanced assistance features that require continuous driver supervision, a driver capable of taking responsibility is present in the vehicle. If the system makes a mistake, that driver is expected to intervene immediately through the steering wheel, brakes or other controls.

In a true robotaxi model, the customer is a passenger; the driving task depends on the vehicle, the remote-support team and the complete operation established by the service provider. If no safety driver is present, it cannot be assumed that the passenger can prevent danger by manually taking control. For this reason, assessing such a service requires looking not only at the vehicle's sensing capability, but also at remote assistance, the call centre, fleet monitoring, towing access and post-incident response procedures.

Being described as “autonomous” does not mean a robotaxi can operate without limits on every road, in every weather condition and in every traffic scenario. Its permitted operating area and conditions are the real determining factors.

Why certification is more complex for vehicles without steering wheels

A significant part of conventional motor-vehicle safety rules was developed on the assumption that a driver is present. Steering wheels, pedals, mirrors, instrument warnings, seating position and the driver's field of vision are all part of that assumption. In a vehicle without a steering wheel and pedals, it must be clearly shown which design, software and operational measures achieve the same safety objectives.

When bodies such as the US National Highway Traffic Safety Administration, NHTSA, request information about certification documentation or a manufacturer's safety assessment, this is sometimes interpreted publicly as an immediate recall. However, an information request or audit query may indicate that the authority wants to review technical documents, compliance reasoning and the process used by the manufacturer. The scope, outcome and any further steps should be followed through official statements; it is not sound to draw final conclusions from early commentary.

Which questions matter in oversight documents?

When a vehicle is declared “compliant with standards,” the more valuable question is how that conclusion was reached. The operator or manufacturer should be able to explain which technical rules it applied, which provisions required alternative-design reasoning, how it defined the test scope, and how software updates are incorporated into safety assessments.

  • Are the vehicle's operational limits clearly defined in terms of service area, speed, weather and road type?

  • What is the redundancy approach for sensing, braking, power supply and communications?

  • Is there a defined procedure for cases where a passenger is stuck in the vehicle, the vehicle is immobilised, or communications are lost?

  • When software changes affect driving behaviour, are testing, logging and rollback mechanisms in place?

  • How are incidents such as crashes, contact events, harsh braking and cancelled trips recorded and investigated?

Safety is shaped by the operation as much as the vehicle

In autonomous fleets, safety is not limited to a vehicle's sensors. The same vehicle may have a different risk profile in a tightly defined area with current maps, a small fleet subject to maintenance discipline and effective remote support than it would in an uncontrolled expansion. That is why beginning in initial cities with limited vehicle numbers, particular operating hours or specific routes can be an understandable operational approach.

Complex intersections, temporary construction layouts, directions from traffic police, closed lanes, dense pedestrian activity and poor visibility are especially critical scenarios for autonomous systems. A good operator does not treat moments when a vehicle cannot proceed as something to hide. It manages them as normal operational situations that require safe stopping, a request for remote help or termination of the trip.

Remote support is not necessarily remote driving

The scope of the term “remote support” used by companies should be clarified. A person in a support centre may communicate with the passenger, suggest a new route to the vehicle, or initiate its safe-stop procedure. Whether the vehicle is actually driven remotely in real time, under what conditions human intervention is activated, and the delay risks associated with that intervention directly affect the service's safety model.

This distinction has practical importance for passengers. The app or in-vehicle screen should clearly show how to request help, how to cancel a trip, what happens if the vehicle stops in a safe location, and which number to call in the event of lost property or a medical emergency.

A first-use checklist for passengers

Before trying a driverless mobility service in Türkiye or abroad, check tangible conditions of use rather than relying on glossy descriptions in an app. As a passenger, you should not be expected to take over driving the vehicle; however, knowing the service limits improves both your safety and your travel experience.

  1. Review the service area: Does the vehicle operate only in certain neighbourhoods, at certain times or in certain weather conditions? Even if your destination is within the limits, is there a restricted area along the route?

  2. Find the emergency-help route: Know the app's live support option, emergency-call button and local emergency number beforehand. In a traffic collision or medical emergency, direct contact with emergency services takes priority.

  3. Wear your seat belt: The absence of a steering wheel does not remove crash risk. Every passenger should wear a belt, and child-seat requirements should be confirmed in advance.

  4. Do not obstruct the vehicle: Do not tamper with doors, sensor areas, screens or any emergency-exit equipment. Luggage and personal items should be placed so they cannot prevent doors from closing.

  5. Document a problem: Report unusual braking, an incorrect route, prolonged immobility or an in-vehicle equipment fault that affects safety to the provider. Where possible, include verifiable details such as trip time and location.

Areas of responsibility for fleet operators and rental companies

A rental or mobility operator using autonomous technology does not transfer its responsibilities simply by purchasing the vehicle. Vehicle suitability, scheduled maintenance, tyre and brake condition, cleaning, data security, customer communication, insurance cover and incident management are core operator duties. Particularly with a software-defined vehicle, the operator must track how updates affect vehicle availability and fleet procedures.

Using the word “autonomous” on its own in contracts is not enough. The features available to a passenger or renter, conditions that may deactivate the system, the support line's responsibility and the reporting process after damage should all be set out clearly. Operators should also train customer-service teams to communicate with anxious passengers and to provide safe guidance when a vehicle is immobilised.

Vehicle buyers and investors: look for evidence, not promises

News about autonomous driving can influence many purchasing decisions, from used-car value to fleet investment. Still, a future software feature, a possible service expansion or a trial in one limited city does not mean that a current vehicle will deliver the same function in every market. A feature being installed in a vehicle also does not mean it is active locally, legal to use or covered from an insurance perspective.

Buyers should separate the name of a feature written in an advertisement from its actual conditions of use. Ask the seller for verifiable information on software version, subscription or licence status, hardware modifications, active warnings, crash history and manufacturer campaigns. In fleet investment, prepare conservative scenarios for vehicle downtime, remote-support costs, the maintenance network and local permit requirements before estimating revenue.

Common misjudgements

The first mistake is treating a federal review automatically as a recall or ban. Reviews may be intended to gather information and assess compliance; the outcome becomes clear only at the end of the official process. The opposite mistake is to assume that every question is insignificant while a review is ongoing. Certification reasoning, field incidents and operational records can be meaningful for public safety.

The second mistake is putting a driverless service and a personal-vehicle feature requiring driver supervision in the same category. Their intended use, chain of responsibility and form of intervention differ. A third mistake is assessing the technology through software alone. Clean sensors, suitable tyres, sound door mechanisms, reliable communications, trained support staff and clear passenger information are integral to the safe operation of an autonomous system.

Takeaway: cautious transparency is the measure of healthy autonomy

Robotaxis may become a valuable option in urban mobility, but trust does not come from a promotional video or a single launch. It comes from defined operating limits, independent and transparent oversight, an operation that learns from incidents, accessible emergency support and honest passenger communication. Removing the steering wheel and pedals does not remove human responsibility; it distributes that responsibility more visibly among the manufacturer, fleet operator, software provider and regulatory authorities.

As a passenger, check the service conditions. As an operator, prepare for the scenarios. As a buyer, separate future-facing promises from the verifiable features available today. The success of autonomous mobility will be determined not simply by whether the vehicle moves, but by how safely and accountably it behaves when the unexpected occurs.

Share
Back to Blog

Related Stories