Key Takeaways
- Research identifies the optimal number of self-driving cars one human can supervise is between five and seven.
- Supervisors overseeing too many vehicles miss critical incidents, while too few can lead to boredom and micromanagement.
- Insights from this study are aiding the development of remote control systems for self-driving vehicles in urban areas.
Research Insights on Supervising Self-Driving Cars
Recent research from Coventry University has answered a crucial question in the transportation industry: how many self-driving cars can one human effectively monitor from a remote location? After years of debate, the study reveals a “Goldilocks zone” where an operator can efficiently supervise five to seven autonomous vehicles.
This research is critical for the future safety and reliability of transport services such as driverless buses, delivery vans, and robotaxis, where a single supervisor must oversee multiple vehicles. Currently, this does not apply to private self-driving cars, which still require a driver to be present in the vehicle.
Published in the journal ‘Computers in Human Behavior,’ the findings emphasize the importance of finding the right supervisory number. Too few vehicles can lead to operator boredom and unnecessary interference, while too many can result in missed critical incidents, posing potential safety risks.
To achieve these conclusions, researchers at Coventry University created a simulator to mimic a future control room. They invited 24 experienced drivers to act as supervisors, monitoring fleets of three, five, seven, and nine self-driving cars navigating a realistic digital replica of Coventry. The participants were instructed to watch the vehicles without intervening, only alerting standby drivers if they noticed something amiss.
The study found that supervising nine vehicles significantly decreased performance, leading supervisors to miss over one-third of critical incidents needing human intervention. While their average response time remained at around 13 seconds, the overload negatively impacted their overall situational awareness and decision-making abilities.
Conversely, when monitoring only three vehicles, supervisors often became bored and started micromanaging, resulting in excessive interference in situations where no action was required. “With three, I felt my attention wandering as there wasn’t so much to focus on,” remarked one participant, further emphasizing the potential risks associated with both extremes.
Ultimately, the study determined that supervising between five and seven self-driving cars struck the perfect balance. Operators reported heightened alertness and faster reaction times without feeling overwhelmed. However, they noted that an influx of text messages from the vehicles could be distracting, suggesting alternative communication methods, like audio alerts, might enhance supervision.
Professor Stewart Birrell, the Director of the Research Centre for Future Transport and Cities, expressed pride in the research’s implications for the safe integration of automated vehicles into everyday environments. “We found that five vehicles are enough to keep operations efficient without overwhelming the human operator,” he said, highlighting the study’s relevance for urban transportation systems where one supervisor may oversee multiple vehicles.
The insights gained are being implemented through the government-funded SCALE project, which aims to establish remote control systems for self-driving vehicles in Solihull. As this research continues to influence emerging transportation frameworks, the findings will play a vital role in shaping the future of connected and automated vehicle systems.
The content above is a summary. For more details, see the source article.
