Electric Cars

Rivian is initiating an exclusive beta program for its cutting-edge point-to-point autonomous driving capability. Only a select group of vehicle owners will be invited to participate, and their involvement will be contingent upon signing a non-disclosure agreement. This strategic move aims to thoroughly test the technology before its broader release.

This initiative, detailed on a recently published support page, targets specific US owners of R1 Gen 2 and R2 vehicles who have subscribed to the Autonomy+ package. The company plans to roll out this feature to the chosen participants in late 2026, with a wider public release not expected until 2027. This phased introduction underscores Rivian's commitment to ensuring safety and reliability in its autonomous systems.

Exclusive Access to Rivian's Autonomous Driving

Rivian's advanced point-to-point driving feature will initially be accessible only through a specialized "Autonomy Early Access Program," requiring invited participants to agree to a non-disclosure clause. This program is exclusively for US-based R1 Gen 2 and R2 owners who subscribe to Autonomy+. The phased rollout is projected for late 2026, with a broader introduction slated for 2027, indicating a deliberate and controlled deployment strategy by Rivian.

To qualify for this invitation-only program, Autonomy+ subscribers must actively engage with Rivian's existing autonomous features and maintain an exemplary safety record. The selection process will also consider geographical distribution and vehicle models, focusing on specific states. First-generation R1 owners and those residing in Canada are not eligible for this initial phase. Participants are required to commit to weekly usage, consistent issue logging, monthly survey completion, and availability for interviews, alongside enabling comprehensive data collection. Membership is subject to monthly eligibility reviews and can be revoked if subscription terms are not met or if the vehicle is sold. In return, testers receive early access to new features and exclusive newsletters.

The Evolving Landscape of Autonomous Technology

The introduction of point-to-point driving marks a significant advancement in Rivian's driver-assistance capabilities, moving beyond the universal hands-free system currently available. While the universal system offers hands-free driving on delineated roads, it lacks the ability to navigate traffic signals or stop signs. The new point-to-point feature aims to provide comprehensive trip management, encompassing a wider range of driving scenarios and complexities.

Rivian's CEO, RJ Scaringe, has characterized the initial version as "Level 2++," necessitating drivers to maintain alertness with eyes on the road, even if hands are off the wheel. The company anticipates achieving hands-off, eyes-off driving by 2027 and Level 4 autonomy by 2028. This timeline represents an adjustment from earlier projections, which had targeted an "eyes-free" system for 2026. The shift underscores the complexities and challenges inherent in developing highly advanced autonomous driving systems. Furthermore, Rivian is set to launch its third-generation autonomy hardware, featuring an in-house RAP1 chip and lidar, by the end of the year. This mirrors a strategy seen with other prominent EV manufacturers, where early beta programs are often accompanied by strict NDAs and meticulous selection processes.

Recent data indicates that Tesla's "Robotaxi" fleet has encountered a recurring issue: minor collisions attributed to remote human intervention rather than the vehicle's autonomous systems. These incidents, where operators take control during perceived system limitations, consistently result in low-speed impacts, bringing into question the reliability of current teleoperation strategies for self-driving vehicles.

Details of Incidents Involving Tesla's Remote-Controlled Robotaxis

In the recent past, the National Highway Traffic Safety Administration (NHTSA) received additional reports detailing two more incidents involving Tesla's "Robotaxis." These occurrences, which transpired in August, were included in NHTSA's September data release, bringing the total number of unique "Robotaxi" incident reports from Tesla to 36. Notably, in all reported cases where a remote operator took over, the vehicle subsequently experienced a collision.

One of the incidents occurred in Dallas, involving a Model Y. The vehicle was stationary on a residential street, positioned between a concrete mixer truck and a parked SUV. A remote operator assumed control, and during a forward movement, the Model Y made contact with the SUV at a speed of approximately 0.40 mph. No passengers were present in the vehicle at the time.

A similar event took place in Austin, where a Model Y, under the control of a remote operator, was navigating through an alley. The vehicle, with a utility pole on one side and a dumpster on the other, collided with the dumpster at around 1 mph. Again, no passengers were onboard. Both incidents resulted solely in property damage and occurred at very low speeds, classified by Tesla as involving a "Remote (Commercial / Test)" driver/operator type.

These two recent crashes align with previous incidents. In July 2025, an Austin-based Tesla, requiring assistance, was remotely operated and accelerated, turning left into a metal fence at 8 mph. Another incident in January 2026, also in Austin, saw a teleoperator guiding the vehicle into a construction barricade at 9 mph after taking control for navigational support. This pattern suggests that when the autonomous system faces challenges or becomes unresponsive, human intervention, despite its intent, has consistently led to minor accidents.

Reflection on the Future of Autonomous Driving and Human Oversight

The consistent pattern of Tesla's "Robotaxis" crashing when under remote human control highlights a critical paradox in the development of autonomous vehicle technology. While the promise of self-driving cars lies in their ability to operate without human error, the need for human intervention in complex or ambiguous situations remains a necessity. However, the data suggests that these remote takeovers, often initiated when the autonomous system is 'stuck' or requires assistance, introduce new vulnerabilities. The inherent challenges of teleoperation, such as latency in communication, limited sensory perception for the remote operator, and a lack of physical feedback from the vehicle, collectively contribute to these unintended collisions. This raises fundamental questions about the optimal balance between autonomous capability and human oversight. As companies like Waymo demonstrate a more cautious approach to remote intervention, limiting human input to guidance rather than direct control, the industry must re-evaluate the design and implementation of teleoperation systems to ensure they genuinely enhance safety and efficiency, rather than inadvertently introducing additional risks.