1–Driver Assistance-–Automated system can sometimes assist the driver in parts of the driving; cruise control, for example. Steering or acceleration/deceleration may be possible, but not both at once.

2–Partial Automation–Automated system can actually conduct some parts of driving while human monitors environment and performs rest of the driving. Here steering and acceleration/deceleration can occur simultaneously.

3–Conditional Automation–Automated system can actually conduct some of driving and monitor environment in some instances, but human must be ready to take back control when the system indicates. Driver can take hands off the wheel and feet off the pedals. The vehicle can conduct all aspects of driving with the expectation that human will respond to request for intervention.

4–High Automation–Automated system can drive and monitor the environment, and the human does not need to take back control, but the vehicle can operate only in certain environments and conditions. For example, vehicles just operate only within certain geo-fenced area.

5–Full Automation–Automated system can perform all driving tasks under all conditions that a human could perform.

 Level 5 is the ultimate goal, and we are not there yet. However, there are many vehicles on the road at varying levels of automation. Because of the growth in this technology, the Insurance Institute for Highway Safety (IIHS) has conducted some tests with vehicles that are at level 2 technology, with adaptive cruise control (ACC) and active lane-keeping. The IIHS wants to establish safety ranking for these technologies the way they test vehicles for other safety parameters. Twenty automakers have stated that by 2025 they will have automatic emergency braking as standard features; knowing that these vehicles perform as advertised is important to consumers.

 Vehicles with level 2 technology help with lane-keeping, speed control and following distance, but the driver must stay in control of the vehicle. Active lane keeping guides the vehicle to stay in the lane, but no more. Multiple tests were conducted with various safety features turned on or off. Five vehicles were tested; one 2017 Mercedes-Benz E-Class with Drive Pilot, a 2017 BMW 5-series with Driving Assistance Plus, a 2018 Tesla model 3 and 2016 Tesla model S with Autopilot, and a 2018 Volvo S90 with Pilot Assist. The tests were conducted as part of a plan to develop a rating system similar to the other IIHS safety ratings for vehicles on the road.

 Tests were conducted both on tracks and on the road. At first vehicles were tested with the autobrake, and not the ACC turned on. When approaching a stationary object the Teslas slowed to reduce the severity of the impact, which is what Tesla claims the technology is designed to do. Other than that, all vehicles tested well when approaching stationary objects, following vehicles that slowed to a stop and then accelerated, and following vehicles that changed lanes to show a stationary object. However, the Teslas braked more smoothly than the other vehicles, and the S90 stopped more abruptly, just 1.1 seconds to avoid impact.

 While the vehicles tested quite well on the tracks for ACC, it was slightly different on the road. The E-class noticed a pick-up truck stopped at a light but then lost track of it so that the driver had to apply the brakes. The Model 3 slowed down for no reason twelve times on the open road, but then resumed speed. The slowdowns were not significant, but could be an issue in heavy traffic. Some of the slowdowns were for tree shadows on the road; others were for oncoming traffic or traffic crossing a road up ahead. However, the Model 3 is the only vehicle that noticed all stopped vehicles in the test. This is the newer camera suite and operating system. The analysis from IIHS indicates that the ACC systems are not truly ready to be used for speed control in all traffic situations on the open road.

 Active lane-keeping was a separate test. The vehicles were tested on known problem areas such as curves and hills without other traffic. The vehicles were also tested on the open road with other traffic. The significance of lane-keeping on curves and hills is whether or not the vehicle will keep from crossing into the other lane on the turns or hills. Only the Model 3 held the lane in all tests. The Model S crossed the center line in one trial; the other vehicles could not consistently stay in their lanes through all tests. The S90 crossed the lane line in eight turns; the 5 series stayed within the lanes only for three trials but would disengage instead of steer outside the lane. Testers noted that some vehicles that relied on road markers were confused by hills. While the Model 3 and the E-Class did ok on the hill trials, the 5-series, Model S and S90 all had difficulty. The 5-series crossed the line regularly causing the drivers to override the system in order to get the vehicle back on track, and sometimes the vehicle disengaged on its own. It failed to stay in the lane on all fourteen trials.

 The Model-S struggled on the hills, swerved back and forth until it determined it was in the right position, and rarely notified the drivers to take over; it just hunted for the lane center. It regularly went into the adjacent lanes or the shoulders. Upon driver intervention, the system disengaged. The S90 did better, disengaging when it crested hills but reengaging when it could again detect the road markings.

 Some vehicles followed a lead vehicle into an exit lane in slow-moving traffic although that is not where the driver intended to go. Apparently when a vehicle is moving too slow to be able to track the lane lines it follows the vehicle in front as a guide, and may follow it in a different direction.

 While these are a small number of tests on only five vehicles, these are common technologies on many vehicles available today, which is why IIHS is working to develop a rating program for such systems.

 While studies show that preventing lane departure crashes could save 8000 lives per year, a system that is not foolproof presents a whole host of different problems. If drivers rely too heavily on technology that isn't flawless, then more accidents are possible. If a driver isn't fully paying attention because he feels the car will not drift over the line, he may not react fast enough to take over from the vehicle if it asks him to take over and disengages. Likewise, if the technology is not accurate enough drivers just won't turn it on at all, making it pointless. Knowing which vehicle's adaptive technology can be relied on to do what it is advertised to do is important to car purchasers. Once IIHS develops a rating system that will be significantly helpful to purchasers to let them know which vehicles truly are safer.