Is ACC actually safer than manual driving in everyday traffic?
Yes, in many everyday situations, ACC improves safety over manual driving. A large naturalistic driving study (where drivers used their own cars in real traffic) found that when ACC and lane keeping systems were active, drivers were significantly less likely to speed and less likely to have a time gap shorter than one second—a major risk factor for rear-end collisions [2]. This means ACC helps maintain safer following distances and speeds without the driver having to constantly adjust.
However, the same study revealed a catch: when drivers overrode the ACC by pressing the gas pedal, they were more likely to speed [2]. So the safety benefit depends on the driver letting the system do its job. Also, different car brands showed different safety performance, suggesting that not all ACC systems are equally effective [2].
How much can ACC reduce crashes?
The crash-reduction potential of ACC is substantial, especially when combined with automatic emergency braking (AEB). A simulation study using real-world motorcycle crash data found that ACC, in its most effective dynamic mode with a 40-meter trigger distance, could prevent 53% of crashes and reduce impact speeds by 4 to 25 km/h [1]. That means more than half of crashes could be avoided entirely, and the rest would be less severe.
Another study on a commercial car with a multi-anticipation ACC (which looks at two cars ahead instead of just one) found that the minimum time-to-collision (TTC) was twice as large when the system was activated, meaning the system gave the driver twice as much time to react to a potential crash [3]. This extra reaction time is a clear safety advantage over manual driving, where drivers often react too late.
When is ACC less safe than manual driving?
ACC is not a magic bullet and can be less safe in certain conditions. Some commercial ACC systems have been shown to have large reaction times and poor 'string stability'—meaning they can amplify small speed changes into larger ones, potentially causing a chain reaction of braking in traffic [3]. This is a problem that manual drivers can sometimes handle better by anticipating traffic flow.
Additionally, ACC systems can be vulnerable to cyberattacks if they rely on vehicle-to-vehicle communication (as in Cooperative ACC). A malicious attack could cause the system to behave dangerously, though researchers are developing machine-learning defenses to detect and block such attacks [4]. Also, ACC is not designed for all scenarios—for example, a new ACC design that specifically handles cut-in vehicles from other lanes was needed to avoid crashes in those situations [5], meaning older ACC systems might not handle sudden lane changes as well as an attentive human driver.
Sources used in this answer
Enhancing motorcycle safety: Quantifying the effects of Autonomous Emergency Braking and Adaptive Cruise Control in crashes reduction.
Motorcycle ACC in dynamic mode with a 40 m trigger distance could prevent 53% of crashes and reduce impact speeds by 4–25 km/h.
Do adaptive cruise control and lane keeping systems make the longitudinal vehicle control safer? Insights into speeding and time gaps shorter than one second from a naturalistic driving study with SAE Level 2 automation
Drivers using ACC and lane keeping were less likely to speed or have a time gap under one second compared to manual driving, but overriding the system increased speeding risk.
Experimental investigation of the multianticipation mechanism in commercial SAE level 2 automated driving vehicles and associated safety impact
A commercial multi-anticipation ACC doubled the minimum time-to-collision compared to when it was not active, but only marginally improved string stability (10%).
Resilient Cooperative Adaptive Cruise Control for Autonomous Vehicles Using Machine Learning
Cooperative ACC (CACC) is vulnerable to malicious V2V attacks that can cause instability, but a machine-learning defense (RACCON) can detect and mitigate such attacks in real time.
A New Adaptive Cruise Control Considering Crash Avoidance for Intelligent Vehicle
A new ACC design with a variable time-to-collision step and switching control successfully avoided crashes with cut-in vehicles from different lanes while maintaining vehicle stability.