Last updated 18-May-2023
It's now 2023 and the 2019 announcement of Deep Rain is maybe finally coming to deliver what it promised, although it still not quite there. That is 4 whole years after they first promised a final solution for eratic windshield wiper performance. The worlds first machine learnt, neural net solution to automatic windscreen wipers. You've got to ask the question though... why?
Before many of us were born, Cadillac in 1950's experimented with rain sensors, GM installed them on convertibles and by 1996 they were adjusting the wiper speed to match conditions. It's a pretty tried and trusted technology that costs about $1 a car and has been on most cars since. Even Tesla installed such a sensor on their cars until the launch of their own autopilot hardware in late 2016.
What Tesla hadn't done back then is work out how to auto control wipers without the sensor. They planned on using the cameras to see if it was raining, but their early attempts took a year to arrive and simply didn't work.
The wipers were erratic, didn't work in the dark, and most owners gave up on them. It wasn't so much of a problem with the MS and MX as the control was on the stalk by the steering wheel, but the minimalist design of the Model 3, and later the Model Y and the 2021 version of the Model S and X meant touching the large screen and finding the option to turn them on.
Tesla even acknowledge the problem as earlier this year they have introduced software updates that enable easier control via the steering wheel controls, but that was partly in repsonse to making them come on automatically when Autopilot is engaged.
As the joke goes, you can tell it doesn't rain much in Fremont.
Tesla have big ideas but often go to market before they have fully resolved them. Autowipers without a rain sensor is just one example. The challenge they faced is how do you know the windscreen needs wiping? The old fashioned sensor worked by detecting moisture on the glass, although it's fair to say that most cars with such a set up also have a sensitivity control allowing fine adjustment. One would think the use of a camera could do the same, ie by seeing the moisture. Except the optics don't work that way, the focal length, focus point and the depth of field of the camera are all set up to see into the distance, and not what's on the glass only a few mm away. Hold your finger in front of your eyes and look to the distance and while you can kind of make out there's a finger there, you can see the distance relatively clearly. This is ideal for autopilot so that small specs of dirt on the camera do not impinge on its capabilities to work, not so good if you're looking for moisture.
So if looking at the screen itself is a challenge, maybe there is an alternative. Why not look at the wider environment for factors to determine if the wipers need to operate, and do this in all conditions and all times of the day. It's no longer a sensor to determine whether the screen is wet, but a system to work out whether its likely to be wet.
There is a compelling argument that says Tesla needs to understand the weather conditions around the car. This is a material input into the AP software as it can have a baring on how far it can see, how fast things may be moving, whether its slippery or hot or foggy etc and use this as a parameter for self driving, but none of those use cases are directly relevant to wiping the windscreen. Tesla and their "must do things differently" approach threw a $1 sensor out the window which could have augmented the systems they had.
Maybe one option is to train the model using a windscreen where the sensor is still in control.
There are some suggestions that Tesla actually tried this, however the failing of the model was that it didn't train on what caused the wipers to sweep, but picked up the sweeping of the wiper itself. In other words, when the car saw the wiper sweep, it realised the windscreen needed swiping, of course in the real world, how does that first sweep occur, and what causes it to realise sweeping is no longer required? Maybe those that experience the frantic windscreen wiping we sometimes see is a legacy of this time. Essentially Tesla need to look at the images that caused the sweep, not the sweep itself, still something that could be achieved.
These are simple steps that could result in a system that in theory would be as good as the sensors, and mixed in manual interventaion, potentially better. There is however still the open question of whether it is even possible. There is a fundamental assumption that looking out the window tells you whether the window needs wiping and we're not sure the case is proven. It's like looking at a million pictures of people to determine whether they have high blood pressure - you would get many right, but would it not be easier to just measure their blood pressure if you could which is what the sensor did? Tesla are making a virtue out of Deep Rain, but in reality it's a problem that didn't really need solving. The same argument extrapolates to every use of AI to solve problems which were, to all intents and purposes, already solved.
There are finally signs that things have improved, however it's still far from perfect. The main issues are dry wiping (the wipers set off when it's not raining), and go crazy when they start (even in light rain they go to max speed). These issues have improved but have not eradicated, but to counter the general improvement, Tesla now turn them on automatically when Autopilot is engaged. If your car suffers from phantom wiper performance, autopilot is now a frustrating experience and while the addition of steering wheel controls has made things slightly better, it is still not really fit for purpose.
By using our site, you acknowledge that you have read and understand our Privacy and Cookie Policy. Your use of the tesla-info website is subject to these policies and terms. All data is provided on a reasonable endeavours basis but errors and omissions may exist. No data should be relied upon as being accurate and additional checks should be made if the information is material to any purchase or use of the car.
Ways you can support tesla-info