AI News Feed
Market watch
Autonomous Driving

Adaptive Cruise Control Still Rests on Two Sensors as Studies Show Driver Overconfidence

Adaptive cruise control still depends on a front radar and a windshield camera, and studies show drivers using it, especially with lane centering, are more likely to speed and take their hands off the wheel.

The hardware arrangement is much the same across brands. In Honda's system, a millimeter-wave radar and a camera measure the gap to the vehicle ahead and how quickly that gap is closing, while the car works the pedals to hold one of four following distances chosen by the driver. The radar, usually mounted behind the grille or the badge, bounces a radio signal off the car in front and times the echo, which yields both distance and closing speed in darkness or fog. Because of where it sits, owners can interfere with it by fitting number-plate frames or stickers, or by bumping into objects while parking.

The camera has an easier life behind the windshield. One of its tasks is reading lane markings so the system knows which vehicle is the right one to follow, which Honda says is also how it recognizes when another driver cuts in. Some automakers add other data sources for smoother behavior: Ford's BlueCruise adds GPS maps, The Drive reported after a ride with Ford engineers.

Hardware performance peaked some years ago. Bosch's long-range radar could already see about 250 meters in 2014, and its successor sees 530, according to Engadget. The more interesting changes have come in software. Honda's Low Speed Follow lets the car stop behind a queue and pull away again when the driver taps a button or the accelerator, making the system genuinely usable in a traffic jam. Ford's Predictive Speed Assist cuts the set speed before a bend and drops it to a new posted limit before the car reaches the sign. GM scans roads with LiDAR for the expansion of Super Cruise, which added 750,000 miles of rural roads. Hyundai built an adaptive cruise control in 2019 that mimics a driver's style using machine learning.

The limits set out in owner manuals are broader than many drivers assume. Toyota says its radar cruise control should not be used in rain, fog or snow, on slippery surfaces, in heavy traffic, on sharp bends or down steep hills. The slippery-surface warning reflects the fact that a computer-held throttle can carry the same speed into standing water, which is how a car hydroplanes; Volvo says its radar's ability to detect the car ahead is reduced significantly in a heavy downpour. Volvo also tells owners not to use adaptive cruise in dense traffic or at junctions, and acknowledges that its sensor can be late to detect a car that suddenly cuts in close ahead. Honda advises using the feature only on relatively straight roads that require little braking.

Volvo's manual states that adaptive cruise control is not a collision avoidance system. It does not brake for people or animals, for small vehicles such as bicycles, or for oncoming, slow-moving or stationary vehicles, including a car sitting stationary in the driver's lane. The system watches moving traffic, so a vehicle that is already stopped, or the back of a traffic jam, may not register at all.

Driver behavior is where the risk compounds. In the IIHS and MIT AgeLab study, 20 volunteers drove cars for a month; half had adaptive cruise alone and half had adaptive cruise plus lane centering. By the end, the lane-centering group was more than three times as likely as in week one to be playing with a phone or the console, and 17 times as likely to have both hands off the wheel. An earlier IIHS study found that drivers with adaptive cruise switched on were 24 percent more likely to speed on highways, and sped by higher margins when they did. All of it surrounds a system that, at bottom, manages a gap and a set speed.