14 ms·

Hard-braking events as indicators of road segment crash risk

370 points by aleyan 8mo ago

harshaw 8mo ago
Nice research. This is fairly well known in insurance circles. Most auto insurers that do telematics consider hard braking the strongest indicator of risk. One of the things that we do at work (Cambridge Mobile Telematics) is build tools to deal with this risk. We have apps that monitor driving and we play a tone to indicate that a hard braking event was detected. Simply letting people know that they had a hard braking event is an effective mechanism for behavior change (other companies have similar tech)
- pavel_lishin 8mo ago
  My mom had a device installed in her car to get a discount on her insurance, and she was always upset at the hard braking thing - whenever she did it, it was because another car was doing something unsafe that she couldn't control, like pulling out in front of her.
  - avidiax 8mo ago
    Some amount of that is inevitable, but there is another level of defensive driving where you anticipate poor behavior and arrange that it won't cause an accident. Have a look at a few dash cam accident videos [1]. There are many maladaptive patterns of behavior, but a frequent one that the average good driver can improve on is limiting speed on two occasions: when approaching a blind spot, and when passing stopped or slow traffic. That second one gets lots of otherwise good drivers. They seem to think that by limiting their speed vs slow/stopped traffic they'd be encouraging people to dart in front of them. Which is somewhat true. But with limited speed, that's an avoidable or less injurious accident. By gunning it past stopped traffic, you make the accident unavoidable and more serious. [1] https://www.youtube.com/@IdiotsInCars1 https://www.youtube.com/@IdiotsInCars1
    - grog454 8mo ago
      The landing page video's first incident is a car coming from behind and from the right, cutting off the filming car. The filming car didn't react at all when instant (but measured) braking would've been safer to start building a distance buffer. One thing HPDE taught me is that most people under brake in dangerous situations because they simply don't know the limit of their vehicle nor the sensitivity range of the brake pedal. The hard braking heuristic makes sense when estimating risk of road segments, but not as a proxy for driver competence.
      - ominous_prime 8mo ago
        It certainly makes sense as a proxy for competence across a diverse population for insurance purposes. You have a baseline of hard braking events that a competent driver may encounter under normal circumstances. If a driver routinely exceeds that number, they are either unable to correctly estimate closing distance and reaction times, which makes them higher risk for causing accidents, or they are driving abnormally aggressively, which also makes them a higher risk for causing accidents. If you consistently put yourself in situations where hard braking is required, it doesn't matter what your skill is, you've reduced your safety margins and an accident is statistically more probable. You said it correctly with "would've been safer to start building a distance buffer", that is the proxy the insurance companies want to use for risk assessment. 35 years without an accident on my record isn't because I'm a magnificent driver, it's because I always try to leave a way out for when something unexpected happens, because the unexpected _does_ happen. The fact that some people may have the skill to drive more aggressively means nothing in the aggregate as far as insurance companies are concerned. If you are skilled enough to drive in that manner, you are skilled enough to avoid it as well. It's simply statistics.
        grog454 8mo ago
        > You said it correctly with "would've been safer to start building a distance buffer", that is the proxy the insurance companies want to use for risk assessment. Then use it? Mandate reaction speed tests or other driving mechanics competency evaluation (not road sign comprehension) and watch insurance margins explode. The driver in my example did poorly and scored top marks in the heuristic.
  - kube-system 8mo ago
    Drivers often believe that their insurance rates should be based solely on whether they follow driving rules, but the risks to insurance are not isolated to this. Someone can follow every rule perfectly, but if they are involved in an accident they incur costs for their insurance company.
    - garaetjjte 8mo ago
      If they are not found at fault then indeed there's no cost for their insurance company?
      - AngryData 8mo ago
        It does cost them just to process the claim because the other guys side is unlikely to completely roll over if there is any chance to reduce the payout. Obviously its not as much as the claims themselves but it also isn't free.
        amanaplanacanal 8mo ago
        Yep. I have a friend that was plowed into from behind while waiting at a red light, twice within a few months time. Two separate intersections. Totaled her car both times. Her insurance rates went up, even though she was clearly not at fault.
        Mawr 8mo ago
        "Not at fault" as far as the law is concerned, is a low bar. If you're one of the last cars in a queue, you should be checking your mirrors for exactly this scenario. Throwing your hands up in the air saying "well the law says I did no wrong!" isn't the way. Unless you like getting into accidents. Also, having two of the same type of accident twice in such a short time is so improbable, that I'm seriously doubting your friend. I'd say it's way more likely she braked too suddenly or has other bad driving habits that make such accidents more likely than that she got this unlucky.
      - 8mo ago
  - infecto 8mo ago
    I hate the warning myself and I use the app the parent is from. I also suspect I am an outlier in not having an accident in 20 years. It’s this obnoxious audio warning that tells me I had a hard breaking and it’s 9/10 because I stopped at a red light that I would not have made on yellow. And then it sends me tips and reminders about reducing hard breaking events and it’s annoying. I know they have done the analysis but it detects moderate hard breaking which is frustrating. One of those things that I am sure in net is positive but perhaps slices of the population it does not benefit.
    - amanaplanacanal 8mo ago
      The lights should be timed so that you don't have to do that if you are driving the posted speed, but I know that's not always true.
      - infecto 8mo ago
        No doubt but I think it’s two fold. Lights stink in that so many are setup not for safety so they rip through the yellow quick as possible. Second, I don’t think of these as hard breaking even but I am sure from a data perspective the cutoff is probably correct. Hard invokes a vision of an emergency brake event. These apps really capture any medium to hard brake event.
      - jeffbee 8mo ago
        It is not possible to time lights that way over any significant distance in a multi-way fabric of traffic.
  - WarmWash 8mo ago
    If you take a seasoned motorcycle rider and put them in one of those dashcam subs, they'll rip their hair out. Most people have near zero defensive driving skill, and view someone pulling out in front of them as "nothing I could have done", when the dashcam shows the offending driver showed 5 signs of pulling out ages before the accident occurred.
    - RupertSalt 8mo ago
      A seasoned motorcycle rider should be unable to rip out any hair, due to safely wearing a helmet!
      - Mawr 8mo ago
        A helmet is priority #100. If your helmet ever comes into play, you've fucked up so bad it'll probably not help you much.
    - organsnyder 8mo ago
      I'm definitely a better driver because of bicycling. You gain new skills when you know that you're going to come out the loser in almost every collision.
      - cucumber3732842 8mo ago
        If you cycle and it doesn't make you a better driver (and pedestrian) you're probably a bad driver and a bad cyclist.
    - doubled112 8mo ago
      At one point in my life I rode a bicycle 40+ km per day. I see things nobody else seems to and I think that has a lot to do with it. I cannot win the collision. Much of being a good driver is just awareness. One time my light turns green, I don't go. As my wife asks what I'm waiting for, a pickup blows the light. We weren't the first car at that light, and years later she still talks about how there's no way I could know. Well, I didn't get us t-boned at 80 so I must have done something right.
- kube-system 8mo ago
  > Simply letting people know that they had a hard braking event is an effective mechanism for behavior change Has this been studied in isolation? Many of the tools that notify upon hard braking also are used to impose financial penalties for doing so... I suspect people may be reacting to the financial incentives.
  - timbaboon 8mo ago
    Yep. We work with CMT and we’ve both done extensive testing on this. I think that often people don’t necessarily know what a hard braking event actually means, or how it’s quantified. Giving that realtime feedback helps close that gap in understanding
  - bluGill 8mo ago
    Does it matter why? The important part is getting people to change behavior. of course if they change such that they don't break hard when needed that is bad, but if the change such that they don't need to break hard in the first place because they slow down in places that are dangerious that is the point.
    - kube-system 8mo ago
      > Does it matter why? The important part is getting people to change behavior. Yeah, if you want to do that, it would be helpful to know whether a financial incentive is required for the effect or not.
  - dghlsakjg 8mo ago
    Anecdotally: I leave the fuel efficiency display as the instrument cluster display on the hybrid that I drive and it significantly changes both my acceleration and braking behavior. There is a minor financial aspect (price of fuel), but I’m far more interested in seeing if I can get a better “green score” at the end of the drive.
  - mpyne 8mo ago
    When people learn to do things by reacting to inputs, they learn much better when the input comes soon after the action/inaction they are trying to train, rather than long after. When you can tie specific acts as a driver to a later financial penalty it helps you learn to avoid the specific acts, otherwise you'd stuck having to figure out in three weeks when the bill comes around what you were doing on the date the insurance statement flagged as a hard stop.
- zahlman 8mo ago
  > We have apps that monitor driving and we play a tone to indicate that a hard braking event was detected. Simply letting people know that they had a hard braking event is an effective mechanism for behavior change (other companies have similar tech) ... How do people not notice that they are braking hard?
  - infecto 8mo ago
    Because their definition of hard is not a slam on the pedal braking. It’s definitely out of the normal stopping but it’s not as hard as you might imagine. I could easily see people not realizing this.
- advisedwang 8mo ago
  Insurance is thinking about hard braking as an indicator of a driver with riskier behaviour. Google is showing that it can also be an indicator of risky road designs. These actually kind of point in opposite directions in terms of the causes of hard braking. The certainly can be used in different ways.
  - alex43578 8mo ago
    A driver who frequents risky roads is a concern to insurers, just as a driver who has risky behaviors. The cause of hard braking isn’t mutually exclusive: bad driving or bad road design.
    - jstanley 8mo ago
      Similarly, a road that is frequently travelled by risky drivers is a risky road!
      - kube-system 8mo ago
        This is part of the reason insurers include zip code in their risk profiles.
  - rogerrogerr 8mo ago
    (Though for an insurer, it’s the same thing - whether you’re risky because you’re a bad driver or because you drive on poorly constructed roads or around other poor drivers is inconsequential to them)
    - mschuster91 8mo ago
      Yeah well fuck insurers. We are supposed to get spied upon by our cars with their blackboxes, by our insurers, by Google, by national security services of various countries... and what do we get in return? Dinged for other people's bad behavior which we cannot reasonably control. Either you follow the car in front of you very closely and get hard braking events, or other people switch lanes in front of you and in the worst case slowing down during lane change, provoking yet another hard braking event. Fuck all of that.
- alwa 8mo ago
  When you modify their braking behavior, is that enough to improve their overall driving behavior? Or do forward collisions and rear-enders make up substantially all of what the driver can control, so training the behaviors to reduce that type of near-miss reduces the driver's overall crash risk? To the point that it's similar to the safest tranche? Is it that hard braking events are broadly indicative of surprises of lots of sorts, and so it happens that the only way to eliminate them all is to develop a full range of defensive driving habits? More Goodhart's Law or Serenity Prayer?
  - toast0 8mo ago
    Regardless of everything else, forward collisions are most likely to have the driver considered at-fault. Seems like reducing those in your insured population would reduce covered losses more than reducing collisions where your insured may not be at fault.
  - Mawr 8mo ago
    > Is it that hard braking events are broadly indicative of surprises of lots of sorts, Of apparent surprises to the driver. And since actual, factual surprises are extremely rare, if a driver is regularly being surprised, they're a bad driver.
- munificent 8mo ago
  How does one not already know that they had a hard braking event? Surely the jamming their foot on the brake pedal and the rapid deceleration would send an even more obvious signal than playing a chime?
  - mecsred 8mo ago
    Obviously people know, but theres no impulse to introspect on why or how. Knowing that someone else knows you had a hard braking event taps in to our social brains to provide a much stronger response to the event. When we know people are watching we're more likely to try and justify our behaviour.
    - redanddead 8mo ago
      that is fascinating
  - infecto 8mo ago
    Have you used one of these apps before? They capture a lot more than emergency stops, what I would classify as the above normal brake effort but not hard braking. Im sure the data exist to set the cutoff but its a lot more than “jam your foot on the pedal braking”. It’s still out of the norm braking for my style of driver but from what I see on the road, people drive aggressively like this. Especially in the US.
    - mhb 8mo ago
      What are these apps or devices? Where can I get one?
      - infecto 8mo ago
        They are provided by your auto insurance company. A potential privacy nightmare, they track your movement. They also can get you a much lower rate.
        8mo ago
drewda 8mo ago
What is the actual use of this? This research team used Google's first-party location data to identify San Jose's Interstate 880/US 101 interchange as a site with statistically extreme amounts of hard braking by Android Auto users. But you don't need machine learning to know that... San Jose Mercury News readers voted that exact location as the worst interchange in the entire Bay Area in a 2018 reader poll [1] It's not a lack of knowledge by Caltrans or Santa Clara County's congestion management agency that is keeping that interchange as-is. Rather, it's the physical constraints of a nearby airport (so no room for flyovers), a nearby river (so probably no tunneling), and surrounding private landowners and train tracks. Leaving aside the specifics of the 880/101 interchange, the Google blog post suggests that they'll use this worst-case scenario on a limited access freeway to inform their future machine-learning analyses of other roads around the country, including ones where presumably there are also pedestrians and cyclists. No doubt some state departments of transportation will line up to buy these new "insights" from Google (forgetting that they actually already buy similar products from TomTom, Inrix, StreetLight, et al.) [2] While I genuinely see the value in data-informed decision making for transportation and urban planning, it's not a lack of data that's causing problems at this particular freeway intersection. This blog post is an underbaked advertisement. [1] https://www.mercurynews.com/2018/04/13/101-880-ranks-as-bay-areas-worst-interchange-roadshow/ https://www.mercurynews.com/2018/04/13/101-880-ranks-as-bay-... [2] https://www.tomtom.com/products/traffic-stats/ https://www.tomtom.com/products/traffic-stats/ https://inrix.com/products/ai-traffic/ https://inrix.com/products/ai-traffic/ https://www.streetlightdata.com/traffic-planning/ https://www.streetlightdata.com/traffic-planning/
- pixl97 8mo ago
  >No doubt some state departments of transportation will line up to buy these new "insights" from Google (forgetting that they actually already buy similar products from TomTom, Inrix, StreetLight, et al.) [2] Google/Apple probably collect a massively larger amount of data than those other companies, putting those other companies at a risk of losing future revenue. Between Google and Apple pretty much every car in the US is monitored.
  - drewda 8mo ago
leetrout 8mo ago
Not surprising but it is nice to have these data streams to explore locations that could potentially be remediated. I think anyone who drives interstates in metro areas would agree cloverleaf interchange are generally terrible with any significant traffic. Add in the general proclivity to drive much higher than the posted speed limit and these become dangerous due to the speed differentials and we've known this for 50 years. "A 1974 study by Hall and Dickinson showed that speed differences contributed to crashes, primarily rear end and lane change collisions" Hall, J. W. and L. V. Dickinson. An Operational Evaluation of Truck Speeds on Interstate Highways, Department of Civil Engineering, University of Maryland, February, 1974.
engelo_b 8mo ago
[dead]
- infecto 8mo ago
  This is definitely pie in the sky but I dream of a future where you have so many autonomous vehicles all the road that we can not only collect this data but also incentivize the slow turning wheels of government to fix it.
  - engelo_b 8mo ago
    [dead]
- pishpash 8mo ago
  Ok but where is the public Maps overlay for this? Is it available?
  - engelo_b 8mo ago
    [dead]
  - jeffbee 8mo ago
    Google is offering this as part of the geospatial platform that they market to governments for huge $$$ so I don't think you are going to get it for free any time soon. Maybe limited access if you have an Earth Engine developer account?
  - joncrane 8mo ago
    Waze has this, and it works off of the same underlying data IIRC. It pops up a "history of accidents" note.
chaps 8mo ago
Like another poster said, this is very well known already. It's one of the reasons why municipalities purchase this data from data brokers.
presidentender 8mo ago
I got one of those dongles from my insurance company that plugged into the ODB2 port and reported my driving habits. I was a bad driver. It would frequently beep at me to let me know that I had braked too hard. I was mystified. "What should I have done differently," I'd think, as I raged at the objective machine that judged me so. The next time my brother came to visit, he called mom. "Oh, and presidentender is a good driver now." I didn't put the pieces together right away, but it turned out that the dongle had actually trained me, like a dog's shock collar. The reason for my too-frequent hard-braking events wasn't speed, although that would be a contributing factor. It was a lack of appropriate following distance. Because I'd follow the drivers in front of me too closely I'd have to brake hard if they did... Or if they drive normally and happened to have a turn coming up. Over the period I had the insurance spy box in my truck I learned without thinking about it to increase my following distance, which meant that riding with me as a passenger was more comfortable and it beeped less often. Of course since I'd been so naughty early during the evaluation they didn't decrease my rates, but I think the training probably did make me statistically less likely to crash.
- xnx 8mo ago
  Thanks for sharing. I'm genuinely impressed to hear someone publicly share a story of growing self awareness and improvement.
- socalgal2 8mo ago
  I'm always surprised at the number of people that follow too closely. This always stuck with me https://www.youtube.com/watch?v=iJFOTSYJrtw&t=466s https://www.youtube.com/watch?v=iJFOTSYJrtw&t=466s
- HoldOnAMinute 8mo ago
  Is there any vehicle that uses it's sensors to make a gentle suggestion about following distance? It's probably the best single thing anyone can do to improve safety. It also reduces wear-and-tear on your car, and increases your fuel economy as a side benefit. Why hasn't gamification of safe driving habits been built directly into the car itself before now?
Someone1234 8mo ago
This type of research is highly valuable but too rare; this is generally because of how we view Road Accidents at a core level: - Road Accidents: "A driver caused this, let's determine who, and find them at fault." - With Air Accidents: "The system caused this, let's determine which elements came together that ultimately lead to this event." The first is essentially simplifying a complex series of events into something black and white. Easy to digest. We'll then keep doing it over and over again because we never changed the circumstances. The second approach is holistic, for example even if the pilot made a mistake, why did they make a mistake, and what can we do to prevent that mistake (e.g. training, culture, etc)? But maybe other elements also played a part like mechanical, software, airport lightning, communications, etc. I bet everyone reading this knows of a road near them that is an accident hotspot and I bet they can explain WHY it is. I certainly do/can, and I see cops with crashed cars there on a weekly basis. Zero changes have been made to the conditions.
- rootusrootus 8mo ago
  An issue of volume, I would guess. If my quickly gathered stats are accurate, there are on the order of 100K commercial flights every day, and 1B drives. So road accidents are expected to be more numerous, and they usually have less impact than an airliner going down. Also - the NTSB does indeed investigate car accidents on occasion, and when they do, they definitely include systemic analysis.
  - trgn 8mo ago
    much easier too investigate and solve though. a few lines of paint here, a bollard there.
- Gigachad 8mo ago
  Because for air travel we pre filter out the morons who are drunk, untrained, scrolling facebook, etc. Any incident left is overwhelmingly more likely a systems issue. For a road incident, the most likely cause was just one driver was violating one or more rules. The only system failure was not canceling their license beforehand.
cbruns 8mo ago
How long until my insurance company can figure out my commute route and adjust my rates based on the collective risk of the segments?
- HoldOnAMinute 8mo ago
  Or it could suggest a less risky route and offer you a discount in exchange for taking that route instead.
  - pishpash 8mo ago
    Why can't Gooapple suggest this today?
    - cbruns 8mo ago
      It seems that they could, but don't.
olliepro 8mo ago
There’s a section of I-15 in Utah’s Salt Lake County which reliably has a crash on weekdays at 6pm. It was unfortunately at a pinch point in the mountains with no good alternate route… very annoying. In a similar way that Google Maps shows eco routes, it’d be fun for them to show “safest” routes which avoid areas with common crashes. (Not always possible, but valuable knowledge when it is.)
- hananova 8mo ago
  That feels like it would cause induced demand for crashes though.
barbazoo 8mo ago
When we worked at a p2p car sharing company it was well understood what a treasure trove that past accelerometer data was as good input to frequency prediction of a claim resulting from a particular rental.
OneOffAsk 8mo ago
Aim to make the road laminar. Every time you hard brake, you're causing the milk jug to glug, making a ripple of entropy as momentum turns to heat from your brakes and those behind you, sometimes in perpetuity. I learned this while doing a 1.5hr daily commute in a Subaru with a clapped out manual transmission. I wanted to conserve energy shifting, but realized I was now participating in large choreographed dance of "smooth" with other drivers who already knew this. There are many of us. And we all glare at the driver blinking their red lights on the interstate indicating that they're loud and proud of introducing turbulence to an otherwise peaceful system.
- nkrisc 8mo ago
  I try to do the same, and do my part to smooth out the wrinkles in traffic. What I would really like in a car is not only my current speed, but the relative speed to the car ahead of me. Given my car has cameras and other sensors for cruise control and other features this ought to already be possible.
- asdff 8mo ago
  https://www.youtube.com/watch?v=Suugn-p5C1M https://www.youtube.com/watch?v=Suugn-p5C1M This is the natural response of tapping the brakes for any slowdown: people naturally over compensating and chain reactions happening behind them. This video shows how stop and go traffic forms and snowballs with no real impetus beyond mis estimated follow distance.
  - alt227 8mo ago
    What this video also shows is that if people pull off more or less at the same time and speed of the car in front then this is a complete non issue. Unfortunately most people are unable to watch a few cars ahead and be ready to pull of when the car in front does, and so we have this situation where each car takes around 5-10 seconds to pull away. multiply this by hundreds or thousands of cars and you end up with a phantom traffic jam. I always pull away at the same speed as the car in front of me and maintain the same distance as when we were stopped. It is very easy to do and completely eliminates traffic build up if multiple people do it at the same time. This is the same reason that we have amber lights on traffic lights, so that the drivers have time to get into gear and start pulling away so that when the light goes green they are imediately travelling through it, causing no excess traffic build up at the lights. Again unfortunately people dont concentrate when they are stopped at lights and so you have the situation where they see the light go green and then proceed to start changing into gear and remove the handbrake. By the time they are moving through the green light, they have already taken 10-20 seconds of green light time, eating well into the time alotted for cars to be travelling across the junction. The only thing which will solve this is driverless cars, meaning that the cars can all talk to each other and move at the same time like a chain. I welcome this advancement to elimante human error in driving and get rid of traffic jams for good.
delichon 8mo ago
I'd love to have a danger heat map displayed on a HUD while driving. Say a default green banner that goes red near a hot spot or even animates near a current hazard. Mostly it could use these same stats, but then be strident if anything unpredictable is detected nearby.
- stevage 8mo ago
  Me too, that's a great idea. Or just incorporated into the satnav, better than getting the warning for an upcoming speed camera.
- itishappy 8mo ago
  Waze used to indicate transient road hazards (e.g. stopped cars, pot holes). It's in Google Maps now too (took them long enough) and I assume others. Static hazards deserve physical signage and/or remediation.
- joncrane 8mo ago
  As I mentioned in another comment, Waze does this. There's a stretch of the Capital Beltway that, if it was on a race track and compressed, would be called "esses." It's totally fine to navigate at 80 MPH with no drama in any mechanically sound, post-1980 car, but it catches mediocre drivers by surprise. Waze throws up a "history of accidents" message whenever I drive through it.
benlivengood 8mo ago
I'm really curious what their data looks like at the various racetracks and circuits. Fun fact; most raceways have accurate street-level indicators (including that they are one-way, but sadly they are not the best racing line) on most online maps, and my car did complain to me in its weekly report about a lot of hard turns, quick acceleration, and hard braking with helpful pins on e.g. Laguna Seca or Thunderhill corners. In theory, the most dangerous turns would probably have higher variance on hard braking data.
roflchoppa 8mo ago
Dude the fucking 101S/808S connector is atrocious on so many levels.
adrianmonk 8mo ago
OK, now that you have this data, give me a "prefer safer routes" option in Google Maps navigation! While you're at it, give me an option to avoid unprotected left turns and to avoid making a left turn across a busy road where cross traffic does not stop. (But only during heavy traffic; it's fine when nobody is on the road.) Not only are these more dangerous, they're also more stressful and they also introduce annoying variation into my travel time.
- stevage 8mo ago
  I think Lyft already does this for their driver navigation.
kazinator 8mo ago
Hard braking could be detected externally; you can tell when vehicles are braking hard from the deceleration and suspension effects, without any surveillance equipment installed in them. That's not gonna be something Google would research, of course, due to next to no alignment with their interests.
rconti 8mo ago
This is a great use of this technology. In aggregate, these hard braking events _do_ tell us about road design issues. They also tell us about problematic drivers, in aggregate. I'll never use one of these dongles, though, because I don't want my every move second-guessed. There's nothing _inherently_ dangerous about isolated hard braking or cornering or acceleration events. It all depends on context. Am I braking hard to avoid an obstacle or mistake by another driver? Is there someone behind me that's likely to rear-end me, or am I in the middle of a highway in the desert? Did I just replace my brake pads and I'm bedding in the new pads? I don't want to have to worry about whether I've used up my invisible quota before the algorithm decides I should be moved into a more expensive insurance bracket.
- rootusrootus 8mo ago
  > There's nothing _inherently_ dangerous about isolated hard braking or cornering or acceleration events. I have to hope that the actuaries at the insurance company are well aware of that. Tuned correctly, the algorithm should not unfairly penalize you. I am reminded of my mother-in-law. She has very few at fault accidents in her decades of driving. And yet she has been involved in a statistically unlikely number of major wrecks. She would say that she is a safe driver, because she is not found at fault. I think it more likely that she is an unsafe driver who creates situations where accidents are more likely to occur but in a way that will not peg her as the underlying cause. Her rates should go up. As should yours, if you are experiencing a very abnormal number of hard-braking events even though you can ascribe every one of them to something outside of your control. Because the implication is that something about your driving habit is increased risk.
  - joncrane 8mo ago
    As far as I understand, many not-at-fault accidents DO make one's rates go up. The rationale being "the places she drives are extra dangerous and puts our client at risk, despite driving properly."
oxag3n 8mo ago
Too bad there's no map with such indicators, I'd definitely use it for my route planning, especially in unknown area. I usually know pretty well dangerous parts if I drive there frequently. In unknown roads/highways I can predict hard bumps/gaps by seeing dark oil spots in the middle of each lane.
jmkd 8mo ago
Against inferential statistics: https://hdl.handle.net/2381/37564 https://hdl.handle.net/2381/37564 PDF download: https://iase-pub.org/ojs/SERJ/article/download/215/119/726 https://iase-pub.org/ojs/SERJ/article/download/215/119/726
BurningFrog 8mo ago
I for one am very glad manual driving is being phased out!
- wiredpancake 8mo ago
  [dead]
luxpir 8mo ago
Wait, what do they teach in North America? Never heard the term "following distance" before now. Sounds misleading. In Britain at least we call it "braking distance" and you're supposed to leave 2 seconds at least between you and the person in front. Count it off a lamp post/sign etc. In certain at-risk areas they use chevrons on the road and signs telling you to keep at least 2 chevrons between you and the car in front. People definitely always get into my braking distance in slower moving traffic, so that happens here too of course. But when things are moving well I likely push the limit and am generally moving faster than most others: going by GPS speed vs speedo, pushing a little into the discretionary and unofficial +10% guidance etc. And weirdly enough I do this for safety and fuel economy. I generally prefer to avoid other vehicles as much as possible in all situations. But I was a motorbike rider in my youth. Once a defensive driver... From that perspective, following distance sounds way more like a gap I want to close up than braking distance does.
- rootusrootus 8mo ago
  > In Britain at least we call it "braking distance" and you're supposed to leave 2 seconds at least between you and the person in front. Count it off a lamp post/sign etc. Indeed, that is the usual definition in the US for following distance. Along with a typical example of how to determine it for yourself. We usually use the term braking distance to describe the distance that would be required to stop the car based on current conditions and speed. This is not necessarily going to be the same as the following distance.
  - luxpir 8mo ago
    Thanks for the reply. So then how is following distance determined? It would have to be further than braking distance to be at least as safe?
    - rootusrootus 8mo ago
      I do not know how they came up with the 2-4 second following distance recommendation. As you point out, this is in line with what the UK recommends as well. Probably it is a compromise between safety and practicality. > It would have to be further than braking distance to be at least as safe? Other way around. Following distance can be less, because the guy you are following cannot stop instantly unless he hits an immovable object or gets into a head-on crash. If he panic stops, then as long as your car performs similarly in braking you just need to have enough distance to allow for your own reaction time. AFAIK braking distance for most cars is around 5 seconds at highway speed. Few people routinely set their following distance that long.
tristor 8mo ago
There are only two conditions under which I have had to hard brake in more than 30 years of driving: 1. I'm on a race track or back road enjoying curves 2. Some asshole did something stupid in front of me. I agree that hard braking is an accurate metric for road segment crash risk, but what I find upsetting is that insurance companies that use vehicle data treat /all/ hard braking equally. In reality, the risk is caused not by every person who hard brakes, but by the first person in a line of cars that hard brakes. More on #2 above is that my observation has been MOST of the time, the braking was COMPLETELY unnecessary. Often the person hard braking that starts the chain has absolutely nothing in front of them to the horizon and is probably on their phone watching TikTok, suddenly looked up and realized they were driving and braked as a spooking reaction. This happens, observably, so often that there are active conspiracies on the Internet that the government hires people to drive like assholes to cause traffic. Obviously that's complete bunk, but my observation here is certainly not unique. Rather than spying on everyone using our vehicle data to charge us ever more money (I've had zero at-fault accidents in nearly 30 years of driving, but my rates only go up), maybe we should enforce attentiveness on the road and start punishing those who are left-lane hogs (causing many lane changes, which are also risky), on their phone, or drunk. It's really obscene these days driving on American roads, it seems like everyone drives markedly worse since the pandemic /and/ enforcement has gone to nothing. The only time I see people get pulled over now is in speedtrap small towns.
JKCalhoun 8mo ago
I considered writing an app you would take on a car ride that would record accelerometer readings indicative of a pothole. Geotag the reading and push is anonymously to a server. Enough events should show clusters where potholes likely exist. You would think cities would love that kind of data.
- jwoy 8mo ago
  This may sound a little cynical but cities often don't like having an inventory of infrastructure problems because it creates an obligation to address it or pay for tire/rim damage if there was prior written notice of a problem. I saw this happen with sidewalk condition data as another example.
djoldman 8mo ago
How happy would HN be if, instead of buying location data as an end run around the 4th amendment, states and municipalities bought traffic data and used it to make safer streets and highways?
killjoywashere 8mo ago
Table 1 and Figure 4: why are the Virginia controlled-access highways different? That population stands out in a way that smells like either a cultural difference or a policy difference. I spend most of my time in California, have lived in SoCal and NorCal, and I spend a fair chunk of time driving around Virginia. My guess is that there's something fishy with the Virginia data being reported. Because if there is anyplace on earth with an insane number of controlled access roads, it's gotta be NVA/DC metro area (or the Tri-Border Area as I like to call it). Also, they need to either update the caption for Figure 4, or move the plots to correspond with the caption. Clearly the Virginia data is on top (or the code is wrong, which seems exceedingly unlikely).
- 1970-01-01 8mo ago
  Speeding in VA just as you would in all other states literally becomes a criminal offense if you're caught. The data is reflective of that.
cabirum 8mo ago
> information from the Android Auto platform Can the quantity and length of tire marks on tarmac be used as a proxy for HBEs?
v4rp1ng 8mo ago
what I thought when I read the article: wait, what? they're collecting this information over my phone from my car!? eventually got me to this: https://arxiv.org/abs/2409.15561 https://arxiv.org/abs/2409.15561 maybe interesting for someone whos privacy focused.
sspehr 8mo ago
Google Maps is definitely one of the technically coolest big tech products imo. It's fascinating what kind of non-obvious insights you can get from the combined user data (E.g. Crowdedness of public places, insights like in the article etc.)
- 4gotunameagain 8mo ago
  The value of said data is tremendous, and could lead to great societal benefits if available for research. So much so that I would argue that google (et al) should be compelled to make the data publicly available. They have enough money, let's make society better.
  - sagarm 8mo ago
    They are making this data publicly available -- after aggregation and anonymization.
    - 4gotunameagain 8mo ago
      A fraction of the data they collect, to keep up appearances.
      - notatoad 8mo ago
        Publicly releasing 100% of the maps data they collect would be the most gigantic privacy violation the world has ever seen.
        4gotunameagain 8mo ago
        Are you seriously suggesting that google is gatekeeping the data because they care about privacy ?
- timmg 8mo ago
  What I find very interesting about it is how much data they have (and are able to collect). I'm personally someone who is less worried about "privacy" for this sort of data. But I know lots of other people feel differently. To me, this is a great example of the "greater good" for data sharing. But it is also a great example of responsible use of data.
zx8080 8mo ago
Nice money they can made selling to insurers the collected data on who drives frequently on those road segments with high risk. There's even no need to get the real names! To anonymously load person's risk coefficient, it's enough to ask to log in using Google account when buying insurance.
HungrMn 8mo ago
[dead]
spacecadet 8mo ago
Defensive driving. Two hands, eyes on the "road", not caring about fools and maintaining safe distances. You have all the time in the world... I Recommend taking your car to a HPDE event, sometimes insurance will discount you for this. In this case, hard braking, is you just not knowing anything about your car or the cars around you... I barely use my brakes, but I also drive manual... Id say learn engine braking but y'all probably driving those stupid over powered electric cars...
- whynotmaybe 8mo ago
  https://donmcminn.com/2020/02/we-judge-ourselves-by-our-intentions-and-others-by-their-behavior/ https://donmcminn.com/2020/02/we-judge-ourselves-by-our-inte...
  - spacecadet 8mo ago
    Nice, but... We live in a world with shared roads... I think a much worse outcome is unfolding as we sell self-driving/assisted EVs to people with little to no real driving experience, while also not upgrading the roads, the laws, and driving education.
1970-01-01 8mo ago
My gut tells me there's also another signal in the data. Soft braking events under conditions that call for acceleration. E.g. driver cannot process a highway merge in real time, so they slow down to have more time to think about what they should do next. Go!
thefringthing 8mo ago
> The core objective was to determine if a high frequency of HBEs causally links to a high rate of crashes. We employed negative binomial (NB) regression models [...]. A negative binomial regression fit on observational data cannot establish that a statistical relationship is causal.
bcrl 8mo ago
Many of the merge lanes in California are insanely short compared to those in the rest of the world. The worst are the ones that have merge immediately before an overpass and exit immediately after where merging and exiting have about the width of the overpass to change lanes. I found those infuriating when I used to visit friends in the Bay area. The pattern where I live is the opposite (long exit lane before the overpass and a long merge lane after) and provides far better margins of safety.

370 points by aleyan 8mo ago

Nice research. This is fairly well known in insurance circles. Most auto insurers that do telematics consider hard braking the strongest indicator of risk. One of the things that we do at work (Cambridge Mobile Telematics) is build tools to deal with this risk. We have apps that monitor driving and we play a tone to indicate that a hard braking event was detected. Simply letting people know that they had a hard braking event is an effective mechanism for behavior change (other companies have similar tech)

pavel_lishin 8mo ago
My mom had a device installed in her car to get a discount on her insurance, and she was always upset at the hard braking thing - whenever she did it, it was because another car was doing something unsafe that she couldn't control, like pulling out in front of her.
- avidiax 8mo ago
  Some amount of that is inevitable, but there is another level of defensive driving where you anticipate poor behavior and arrange that it won't cause an accident. Have a look at a few dash cam accident videos [1]. There are many maladaptive patterns of behavior, but a frequent one that the average good driver can improve on is limiting speed on two occasions: when approaching a blind spot, and when passing stopped or slow traffic. That second one gets lots of otherwise good drivers. They seem to think that by limiting their speed vs slow/stopped traffic they'd be encouraging people to dart in front of them. Which is somewhat true. But with limited speed, that's an avoidable or less injurious accident. By gunning it past stopped traffic, you make the accident unavoidable and more serious. [1] https://www.youtube.com/@IdiotsInCars1 https://www.youtube.com/@IdiotsInCars1
  - grog454 8mo ago
    The landing page video's first incident is a car coming from behind and from the right, cutting off the filming car. The filming car didn't react at all when instant (but measured) braking would've been safer to start building a distance buffer. One thing HPDE taught me is that most people under brake in dangerous situations because they simply don't know the limit of their vehicle nor the sensitivity range of the brake pedal. The hard braking heuristic makes sense when estimating risk of road segments, but not as a proxy for driver competence.
    - ominous_prime 8mo ago
      It certainly makes sense as a proxy for competence across a diverse population for insurance purposes. You have a baseline of hard braking events that a competent driver may encounter under normal circumstances. If a driver routinely exceeds that number, they are either unable to correctly estimate closing distance and reaction times, which makes them higher risk for causing accidents, or they are driving abnormally aggressively, which also makes them a higher risk for causing accidents. If you consistently put yourself in situations where hard braking is required, it doesn't matter what your skill is, you've reduced your safety margins and an accident is statistically more probable. You said it correctly with "would've been safer to start building a distance buffer", that is the proxy the insurance companies want to use for risk assessment. 35 years without an accident on my record isn't because I'm a magnificent driver, it's because I always try to leave a way out for when something unexpected happens, because the unexpected _does_ happen. The fact that some people may have the skill to drive more aggressively means nothing in the aggregate as far as insurance companies are concerned. If you are skilled enough to drive in that manner, you are skilled enough to avoid it as well. It's simply statistics.
      - grog454 8mo ago
        > You said it correctly with "would've been safer to start building a distance buffer", that is the proxy the insurance companies want to use for risk assessment. Then use it? Mandate reaction speed tests or other driving mechanics competency evaluation (not road sign comprehension) and watch insurance margins explode. The driver in my example did poorly and scored top marks in the heuristic.
- kube-system 8mo ago
  Drivers often believe that their insurance rates should be based solely on whether they follow driving rules, but the risks to insurance are not isolated to this. Someone can follow every rule perfectly, but if they are involved in an accident they incur costs for their insurance company.
  - garaetjjte 8mo ago
    If they are not found at fault then indeed there's no cost for their insurance company?
    - AngryData 8mo ago
      It does cost them just to process the claim because the other guys side is unlikely to completely roll over if there is any chance to reduce the payout. Obviously its not as much as the claims themselves but it also isn't free.
      - amanaplanacanal 8mo ago
        Yep. I have a friend that was plowed into from behind while waiting at a red light, twice within a few months time. Two separate intersections. Totaled her car both times. Her insurance rates went up, even though she was clearly not at fault.
        Mawr 8mo ago
        "Not at fault" as far as the law is concerned, is a low bar. If you're one of the last cars in a queue, you should be checking your mirrors for exactly this scenario. Throwing your hands up in the air saying "well the law says I did no wrong!" isn't the way. Unless you like getting into accidents. Also, having two of the same type of accident twice in such a short time is so improbable, that I'm seriously doubting your friend. I'd say it's way more likely she braked too suddenly or has other bad driving habits that make such accidents more likely than that she got this unlucky.
    - 8mo ago
- infecto 8mo ago
  I hate the warning myself and I use the app the parent is from. I also suspect I am an outlier in not having an accident in 20 years. It’s this obnoxious audio warning that tells me I had a hard breaking and it’s 9/10 because I stopped at a red light that I would not have made on yellow. And then it sends me tips and reminders about reducing hard breaking events and it’s annoying. I know they have done the analysis but it detects moderate hard breaking which is frustrating. One of those things that I am sure in net is positive but perhaps slices of the population it does not benefit.
  - amanaplanacanal 8mo ago
    The lights should be timed so that you don't have to do that if you are driving the posted speed, but I know that's not always true.
    - infecto 8mo ago
      No doubt but I think it’s two fold. Lights stink in that so many are setup not for safety so they rip through the yellow quick as possible. Second, I don’t think of these as hard breaking even but I am sure from a data perspective the cutoff is probably correct. Hard invokes a vision of an emergency brake event. These apps really capture any medium to hard brake event.
    - jeffbee 8mo ago
      It is not possible to time lights that way over any significant distance in a multi-way fabric of traffic.
- WarmWash 8mo ago
  If you take a seasoned motorcycle rider and put them in one of those dashcam subs, they'll rip their hair out. Most people have near zero defensive driving skill, and view someone pulling out in front of them as "nothing I could have done", when the dashcam shows the offending driver showed 5 signs of pulling out ages before the accident occurred.
  - RupertSalt 8mo ago
    A seasoned motorcycle rider should be unable to rip out any hair, due to safely wearing a helmet!
    - Mawr 8mo ago
      A helmet is priority #100. If your helmet ever comes into play, you've fucked up so bad it'll probably not help you much.
  - organsnyder 8mo ago
    I'm definitely a better driver because of bicycling. You gain new skills when you know that you're going to come out the loser in almost every collision.
    - cucumber3732842 8mo ago
      If you cycle and it doesn't make you a better driver (and pedestrian) you're probably a bad driver and a bad cyclist.
  - doubled112 8mo ago
    At one point in my life I rode a bicycle 40+ km per day. I see things nobody else seems to and I think that has a lot to do with it. I cannot win the collision. Much of being a good driver is just awareness. One time my light turns green, I don't go. As my wife asks what I'm waiting for, a pickup blows the light. We weren't the first car at that light, and years later she still talks about how there's no way I could know. Well, I didn't get us t-boned at 80 so I must have done something right.
kube-system 8mo ago
> Simply letting people know that they had a hard braking event is an effective mechanism for behavior change Has this been studied in isolation? Many of the tools that notify upon hard braking also are used to impose financial penalties for doing so... I suspect people may be reacting to the financial incentives.
- timbaboon 8mo ago
  Yep. We work with CMT and we’ve both done extensive testing on this. I think that often people don’t necessarily know what a hard braking event actually means, or how it’s quantified. Giving that realtime feedback helps close that gap in understanding
- bluGill 8mo ago
  Does it matter why? The important part is getting people to change behavior. of course if they change such that they don't break hard when needed that is bad, but if the change such that they don't need to break hard in the first place because they slow down in places that are dangerious that is the point.
  - kube-system 8mo ago
    > Does it matter why? The important part is getting people to change behavior. Yeah, if you want to do that, it would be helpful to know whether a financial incentive is required for the effect or not.
- dghlsakjg 8mo ago
  Anecdotally: I leave the fuel efficiency display as the instrument cluster display on the hybrid that I drive and it significantly changes both my acceleration and braking behavior. There is a minor financial aspect (price of fuel), but I’m far more interested in seeing if I can get a better “green score” at the end of the drive.
- mpyne 8mo ago
  When people learn to do things by reacting to inputs, they learn much better when the input comes soon after the action/inaction they are trying to train, rather than long after. When you can tie specific acts as a driver to a later financial penalty it helps you learn to avoid the specific acts, otherwise you'd stuck having to figure out in three weeks when the bill comes around what you were doing on the date the insurance statement flagged as a hard stop.
zahlman 8mo ago
> We have apps that monitor driving and we play a tone to indicate that a hard braking event was detected. Simply letting people know that they had a hard braking event is an effective mechanism for behavior change (other companies have similar tech) ... How do people not notice that they are braking hard?
- infecto 8mo ago
  Because their definition of hard is not a slam on the pedal braking. It’s definitely out of the normal stopping but it’s not as hard as you might imagine. I could easily see people not realizing this.
advisedwang 8mo ago
Insurance is thinking about hard braking as an indicator of a driver with riskier behaviour. Google is showing that it can also be an indicator of risky road designs. These actually kind of point in opposite directions in terms of the causes of hard braking. The certainly can be used in different ways.
- alex43578 8mo ago
  A driver who frequents risky roads is a concern to insurers, just as a driver who has risky behaviors. The cause of hard braking isn’t mutually exclusive: bad driving or bad road design.
  - jstanley 8mo ago
    Similarly, a road that is frequently travelled by risky drivers is a risky road!
    - kube-system 8mo ago
      This is part of the reason insurers include zip code in their risk profiles.
- rogerrogerr 8mo ago
  (Though for an insurer, it’s the same thing - whether you’re risky because you’re a bad driver or because you drive on poorly constructed roads or around other poor drivers is inconsequential to them)
  - mschuster91 8mo ago
    Yeah well fuck insurers. We are supposed to get spied upon by our cars with their blackboxes, by our insurers, by Google, by national security services of various countries... and what do we get in return? Dinged for other people's bad behavior which we cannot reasonably control. Either you follow the car in front of you very closely and get hard braking events, or other people switch lanes in front of you and in the worst case slowing down during lane change, provoking yet another hard braking event. Fuck all of that.
alwa 8mo ago
When you modify their braking behavior, is that enough to improve their overall driving behavior? Or do forward collisions and rear-enders make up substantially all of what the driver can control, so training the behaviors to reduce that type of near-miss reduces the driver's overall crash risk? To the point that it's similar to the safest tranche? Is it that hard braking events are broadly indicative of surprises of lots of sorts, and so it happens that the only way to eliminate them all is to develop a full range of defensive driving habits? More Goodhart's Law or Serenity Prayer?
- toast0 8mo ago
  Regardless of everything else, forward collisions are most likely to have the driver considered at-fault. Seems like reducing those in your insured population would reduce covered losses more than reducing collisions where your insured may not be at fault.
- Mawr 8mo ago
  > Is it that hard braking events are broadly indicative of surprises of lots of sorts, Of apparent surprises to the driver. And since actual, factual surprises are extremely rare, if a driver is regularly being surprised, they're a bad driver.
munificent 8mo ago
How does one not already know that they had a hard braking event? Surely the jamming their foot on the brake pedal and the rapid deceleration would send an even more obvious signal than playing a chime?
- mecsred 8mo ago
  Obviously people know, but theres no impulse to introspect on why or how. Knowing that someone else knows you had a hard braking event taps in to our social brains to provide a much stronger response to the event. When we know people are watching we're more likely to try and justify our behaviour.
  - redanddead 8mo ago
    that is fascinating
- infecto 8mo ago
  Have you used one of these apps before? They capture a lot more than emergency stops, what I would classify as the above normal brake effort but not hard braking. Im sure the data exist to set the cutoff but its a lot more than “jam your foot on the pedal braking”. It’s still out of the norm braking for my style of driver but from what I see on the road, people drive aggressively like this. Especially in the US.
  - mhb 8mo ago
    What are these apps or devices? Where can I get one?
    - infecto 8mo ago
      They are provided by your auto insurance company. A potential privacy nightmare, they track your movement. They also can get you a much lower rate.
      - 8mo ago

drewda 8mo ago

What is the actual use of this? This research team used Google's first-party location data to identify San Jose's Interstate 880/US 101 interchange as a site with statistically extreme amounts of hard braking by Android Auto users. But you don't need machine learning to know that... San Jose Mercury News readers voted that exact location as the worst interchange in the entire Bay Area in a 2018 reader poll [1] It's not a lack of knowledge by Caltrans or Santa Clara County's congestion management agency that is keeping that interchange as-is. Rather, it's the physical constraints of a nearby airport (so no room for flyovers), a nearby river (so probably no tunneling), and surrounding private landowners and train tracks. Leaving aside the specifics of the 880/101 interchange, the Google blog post suggests that they'll use this worst-case scenario on a limited access freeway to inform their future machine-learning analyses of other roads around the country, including ones where presumably there are also pedestrians and cyclists. No doubt some state departments of transportation will line up to buy these new "insights" from Google (forgetting that they actually already buy similar products from TomTom, Inrix, StreetLight, et al.) [2] While I genuinely see the value in data-informed decision making for transportation and urban planning, it's not a lack of data that's causing problems at this particular freeway intersection. This blog post is an underbaked advertisement. [1] https://www.mercurynews.com/2018/04/13/101-880-ranks-as-bay-areas-worst-interchange-roadshow/ https://www.mercurynews.com/2018/04/13/101-880-ranks-as-bay-... [2] https://www.tomtom.com/products/traffic-stats/ https://www.tomtom.com/products/traffic-stats/ https://inrix.com/products/ai-traffic/ https://inrix.com/products/ai-traffic/ https://www.streetlightdata.com/traffic-planning/ https://www.streetlightdata.com/traffic-planning/

pixl97 8mo ago
>No doubt some state departments of transportation will line up to buy these new "insights" from Google (forgetting that they actually already buy similar products from TomTom, Inrix, StreetLight, et al.) [2] Google/Apple probably collect a massively larger amount of data than those other companies, putting those other companies at a risk of losing future revenue. Between Google and Apple pretty much every car in the US is monitored.
- drewda 8mo ago

leetrout 8mo ago

Not surprising but it is nice to have these data streams to explore locations that could potentially be remediated. I think anyone who drives interstates in metro areas would agree cloverleaf interchange are generally terrible with any significant traffic. Add in the general proclivity to drive much higher than the posted speed limit and these become dangerous due to the speed differentials and we've known this for 50 years. "A 1974 study by Hall and Dickinson showed that speed differences contributed to crashes, primarily rear end and lane change collisions" Hall, J. W. and L. V. Dickinson. An Operational Evaluation of Truck Speeds on Interstate Highways, Department of Civil Engineering, University of Maryland, February, 1974.

engelo_b 8mo ago

[dead]

infecto 8mo ago
This is definitely pie in the sky but I dream of a future where you have so many autonomous vehicles all the road that we can not only collect this data but also incentivize the slow turning wheels of government to fix it.
- engelo_b 8mo ago
  [dead]
pishpash 8mo ago
Ok but where is the public Maps overlay for this? Is it available?
- engelo_b 8mo ago
  [dead]
- jeffbee 8mo ago
  Google is offering this as part of the geospatial platform that they market to governments for huge $$$ so I don't think you are going to get it for free any time soon. Maybe limited access if you have an Earth Engine developer account?
- joncrane 8mo ago
  Waze has this, and it works off of the same underlying data IIRC. It pops up a "history of accidents" note.

chaps 8mo ago

Like another poster said, this is very well known already. It's one of the reasons why municipalities purchase this data from data brokers.

presidentender 8mo ago

I got one of those dongles from my insurance company that plugged into the ODB2 port and reported my driving habits. I was a bad driver. It would frequently beep at me to let me know that I had braked too hard. I was mystified. "What should I have done differently," I'd think, as I raged at the objective machine that judged me so. The next time my brother came to visit, he called mom. "Oh, and presidentender is a good driver now." I didn't put the pieces together right away, but it turned out that the dongle had actually trained me, like a dog's shock collar. The reason for my too-frequent hard-braking events wasn't speed, although that would be a contributing factor. It was a lack of appropriate following distance. Because I'd follow the drivers in front of me too closely I'd have to brake hard if they did... Or if they drive normally and happened to have a turn coming up. Over the period I had the insurance spy box in my truck I learned without thinking about it to increase my following distance, which meant that riding with me as a passenger was more comfortable and it beeped less often. Of course since I'd been so naughty early during the evaluation they didn't decrease my rates, but I think the training probably did make me statistically less likely to crash.

xnx 8mo ago
Thanks for sharing. I'm genuinely impressed to hear someone publicly share a story of growing self awareness and improvement.
socalgal2 8mo ago
I'm always surprised at the number of people that follow too closely. This always stuck with me https://www.youtube.com/watch?v=iJFOTSYJrtw&t=466s https://www.youtube.com/watch?v=iJFOTSYJrtw&t=466s
HoldOnAMinute 8mo ago
Is there any vehicle that uses it's sensors to make a gentle suggestion about following distance? It's probably the best single thing anyone can do to improve safety. It also reduces wear-and-tear on your car, and increases your fuel economy as a side benefit. Why hasn't gamification of safe driving habits been built directly into the car itself before now?

Someone1234 8mo ago

This type of research is highly valuable but too rare; this is generally because of how we view Road Accidents at a core level: - Road Accidents: "A driver caused this, let's determine who, and find them at fault." - With Air Accidents: "The system caused this, let's determine which elements came together that ultimately lead to this event." The first is essentially simplifying a complex series of events into something black and white. Easy to digest. We'll then keep doing it over and over again because we never changed the circumstances. The second approach is holistic, for example even if the pilot made a mistake, why did they make a mistake, and what can we do to prevent that mistake (e.g. training, culture, etc)? But maybe other elements also played a part like mechanical, software, airport lightning, communications, etc. I bet everyone reading this knows of a road near them that is an accident hotspot and I bet they can explain WHY it is. I certainly do/can, and I see cops with crashed cars there on a weekly basis. Zero changes have been made to the conditions.

rootusrootus 8mo ago
An issue of volume, I would guess. If my quickly gathered stats are accurate, there are on the order of 100K commercial flights every day, and 1B drives. So road accidents are expected to be more numerous, and they usually have less impact than an airliner going down. Also - the NTSB does indeed investigate car accidents on occasion, and when they do, they definitely include systemic analysis.
- trgn 8mo ago
  much easier too investigate and solve though. a few lines of paint here, a bollard there.
Gigachad 8mo ago
Because for air travel we pre filter out the morons who are drunk, untrained, scrolling facebook, etc. Any incident left is overwhelmingly more likely a systems issue. For a road incident, the most likely cause was just one driver was violating one or more rules. The only system failure was not canceling their license beforehand.

cbruns 8mo ago

How long until my insurance company can figure out my commute route and adjust my rates based on the collective risk of the segments?

HoldOnAMinute 8mo ago
Or it could suggest a less risky route and offer you a discount in exchange for taking that route instead.
- pishpash 8mo ago
  Why can't Gooapple suggest this today?
  - cbruns 8mo ago
    It seems that they could, but don't.

olliepro 8mo ago

There’s a section of I-15 in Utah’s Salt Lake County which reliably has a crash on weekdays at 6pm. It was unfortunately at a pinch point in the mountains with no good alternate route… very annoying. In a similar way that Google Maps shows eco routes, it’d be fun for them to show “safest” routes which avoid areas with common crashes. (Not always possible, but valuable knowledge when it is.)

hananova 8mo ago
That feels like it would cause induced demand for crashes though.

barbazoo 8mo ago

When we worked at a p2p car sharing company it was well understood what a treasure trove that past accelerometer data was as good input to frequency prediction of a claim resulting from a particular rental.

OneOffAsk 8mo ago

Aim to make the road laminar. Every time you hard brake, you're causing the milk jug to glug, making a ripple of entropy as momentum turns to heat from your brakes and those behind you, sometimes in perpetuity. I learned this while doing a 1.5hr daily commute in a Subaru with a clapped out manual transmission. I wanted to conserve energy shifting, but realized I was now participating in large choreographed dance of "smooth" with other drivers who already knew this. There are many of us. And we all glare at the driver blinking their red lights on the interstate indicating that they're loud and proud of introducing turbulence to an otherwise peaceful system.

nkrisc 8mo ago
I try to do the same, and do my part to smooth out the wrinkles in traffic. What I would really like in a car is not only my current speed, but the relative speed to the car ahead of me. Given my car has cameras and other sensors for cruise control and other features this ought to already be possible.
asdff 8mo ago
https://www.youtube.com/watch?v=Suugn-p5C1M https://www.youtube.com/watch?v=Suugn-p5C1M This is the natural response of tapping the brakes for any slowdown: people naturally over compensating and chain reactions happening behind them. This video shows how stop and go traffic forms and snowballs with no real impetus beyond mis estimated follow distance.
- alt227 8mo ago
  What this video also shows is that if people pull off more or less at the same time and speed of the car in front then this is a complete non issue. Unfortunately most people are unable to watch a few cars ahead and be ready to pull of when the car in front does, and so we have this situation where each car takes around 5-10 seconds to pull away. multiply this by hundreds or thousands of cars and you end up with a phantom traffic jam. I always pull away at the same speed as the car in front of me and maintain the same distance as when we were stopped. It is very easy to do and completely eliminates traffic build up if multiple people do it at the same time. This is the same reason that we have amber lights on traffic lights, so that the drivers have time to get into gear and start pulling away so that when the light goes green they are imediately travelling through it, causing no excess traffic build up at the lights. Again unfortunately people dont concentrate when they are stopped at lights and so you have the situation where they see the light go green and then proceed to start changing into gear and remove the handbrake. By the time they are moving through the green light, they have already taken 10-20 seconds of green light time, eating well into the time alotted for cars to be travelling across the junction. The only thing which will solve this is driverless cars, meaning that the cars can all talk to each other and move at the same time like a chain. I welcome this advancement to elimante human error in driving and get rid of traffic jams for good.

delichon 8mo ago

I'd love to have a danger heat map displayed on a HUD while driving. Say a default green banner that goes red near a hot spot or even animates near a current hazard. Mostly it could use these same stats, but then be strident if anything unpredictable is detected nearby.

stevage 8mo ago
Me too, that's a great idea. Or just incorporated into the satnav, better than getting the warning for an upcoming speed camera.
itishappy 8mo ago
Waze used to indicate transient road hazards (e.g. stopped cars, pot holes). It's in Google Maps now too (took them long enough) and I assume others. Static hazards deserve physical signage and/or remediation.
joncrane 8mo ago
As I mentioned in another comment, Waze does this. There's a stretch of the Capital Beltway that, if it was on a race track and compressed, would be called "esses." It's totally fine to navigate at 80 MPH with no drama in any mechanically sound, post-1980 car, but it catches mediocre drivers by surprise. Waze throws up a "history of accidents" message whenever I drive through it.

benlivengood 8mo ago

I'm really curious what their data looks like at the various racetracks and circuits. Fun fact; most raceways have accurate street-level indicators (including that they are one-way, but sadly they are not the best racing line) on most online maps, and my car did complain to me in its weekly report about a lot of hard turns, quick acceleration, and hard braking with helpful pins on e.g. Laguna Seca or Thunderhill corners. In theory, the most dangerous turns would probably have higher variance on hard braking data.

roflchoppa 8mo ago

Dude the fucking 101S/808S connector is atrocious on so many levels.

adrianmonk 8mo ago

OK, now that you have this data, give me a "prefer safer routes" option in Google Maps navigation! While you're at it, give me an option to avoid unprotected left turns and to avoid making a left turn across a busy road where cross traffic does not stop. (But only during heavy traffic; it's fine when nobody is on the road.) Not only are these more dangerous, they're also more stressful and they also introduce annoying variation into my travel time.

stevage 8mo ago
I think Lyft already does this for their driver navigation.

kazinator 8mo ago

Hard braking could be detected externally; you can tell when vehicles are braking hard from the deceleration and suspension effects, without any surveillance equipment installed in them. That's not gonna be something Google would research, of course, due to next to no alignment with their interests.

rconti 8mo ago

This is a great use of this technology. In aggregate, these hard braking events _do_ tell us about road design issues. They also tell us about problematic drivers, in aggregate. I'll never use one of these dongles, though, because I don't want my every move second-guessed. There's nothing _inherently_ dangerous about isolated hard braking or cornering or acceleration events. It all depends on context. Am I braking hard to avoid an obstacle or mistake by another driver? Is there someone behind me that's likely to rear-end me, or am I in the middle of a highway in the desert? Did I just replace my brake pads and I'm bedding in the new pads? I don't want to have to worry about whether I've used up my invisible quota before the algorithm decides I should be moved into a more expensive insurance bracket.

rootusrootus 8mo ago
> There's nothing _inherently_ dangerous about isolated hard braking or cornering or acceleration events. I have to hope that the actuaries at the insurance company are well aware of that. Tuned correctly, the algorithm should not unfairly penalize you. I am reminded of my mother-in-law. She has very few at fault accidents in her decades of driving. And yet she has been involved in a statistically unlikely number of major wrecks. She would say that she is a safe driver, because she is not found at fault. I think it more likely that she is an unsafe driver who creates situations where accidents are more likely to occur but in a way that will not peg her as the underlying cause. Her rates should go up. As should yours, if you are experiencing a very abnormal number of hard-braking events even though you can ascribe every one of them to something outside of your control. Because the implication is that something about your driving habit is increased risk.
- joncrane 8mo ago
  As far as I understand, many not-at-fault accidents DO make one's rates go up. The rationale being "the places she drives are extra dangerous and puts our client at risk, despite driving properly."

oxag3n 8mo ago

Too bad there's no map with such indicators, I'd definitely use it for my route planning, especially in unknown area. I usually know pretty well dangerous parts if I drive there frequently. In unknown roads/highways I can predict hard bumps/gaps by seeing dark oil spots in the middle of each lane.

jmkd 8mo ago

Against inferential statistics: https://hdl.handle.net/2381/37564 https://hdl.handle.net/2381/37564 PDF download: https://iase-pub.org/ojs/SERJ/article/download/215/119/726 https://iase-pub.org/ojs/SERJ/article/download/215/119/726

BurningFrog 8mo ago

I for one am very glad manual driving is being phased out!

wiredpancake 8mo ago
[dead]

luxpir 8mo ago

Wait, what do they teach in North America? Never heard the term "following distance" before now. Sounds misleading. In Britain at least we call it "braking distance" and you're supposed to leave 2 seconds at least between you and the person in front. Count it off a lamp post/sign etc. In certain at-risk areas they use chevrons on the road and signs telling you to keep at least 2 chevrons between you and the car in front. People definitely always get into my braking distance in slower moving traffic, so that happens here too of course. But when things are moving well I likely push the limit and am generally moving faster than most others: going by GPS speed vs speedo, pushing a little into the discretionary and unofficial +10% guidance etc. And weirdly enough I do this for safety and fuel economy. I generally prefer to avoid other vehicles as much as possible in all situations. But I was a motorbike rider in my youth. Once a defensive driver... From that perspective, following distance sounds way more like a gap I want to close up than braking distance does.

rootusrootus 8mo ago
> In Britain at least we call it "braking distance" and you're supposed to leave 2 seconds at least between you and the person in front. Count it off a lamp post/sign etc. Indeed, that is the usual definition in the US for following distance. Along with a typical example of how to determine it for yourself. We usually use the term braking distance to describe the distance that would be required to stop the car based on current conditions and speed. This is not necessarily going to be the same as the following distance.
- luxpir 8mo ago
  Thanks for the reply. So then how is following distance determined? It would have to be further than braking distance to be at least as safe?
  - rootusrootus 8mo ago
    I do not know how they came up with the 2-4 second following distance recommendation. As you point out, this is in line with what the UK recommends as well. Probably it is a compromise between safety and practicality. > It would have to be further than braking distance to be at least as safe? Other way around. Following distance can be less, because the guy you are following cannot stop instantly unless he hits an immovable object or gets into a head-on crash. If he panic stops, then as long as your car performs similarly in braking you just need to have enough distance to allow for your own reaction time. AFAIK braking distance for most cars is around 5 seconds at highway speed. Few people routinely set their following distance that long.

tristor 8mo ago

There are only two conditions under which I have had to hard brake in more than 30 years of driving: 1. I'm on a race track or back road enjoying curves 2. Some asshole did something stupid in front of me. I agree that hard braking is an accurate metric for road segment crash risk, but what I find upsetting is that insurance companies that use vehicle data treat /all/ hard braking equally. In reality, the risk is caused not by every person who hard brakes, but by the first person in a line of cars that hard brakes. More on #2 above is that my observation has been MOST of the time, the braking was COMPLETELY unnecessary. Often the person hard braking that starts the chain has absolutely nothing in front of them to the horizon and is probably on their phone watching TikTok, suddenly looked up and realized they were driving and braked as a spooking reaction. This happens, observably, so often that there are active conspiracies on the Internet that the government hires people to drive like assholes to cause traffic. Obviously that's complete bunk, but my observation here is certainly not unique. Rather than spying on everyone using our vehicle data to charge us ever more money (I've had zero at-fault accidents in nearly 30 years of driving, but my rates only go up), maybe we should enforce attentiveness on the road and start punishing those who are left-lane hogs (causing many lane changes, which are also risky), on their phone, or drunk. It's really obscene these days driving on American roads, it seems like everyone drives markedly worse since the pandemic /and/ enforcement has gone to nothing. The only time I see people get pulled over now is in speedtrap small towns.

JKCalhoun 8mo ago

I considered writing an app you would take on a car ride that would record accelerometer readings indicative of a pothole. Geotag the reading and push is anonymously to a server. Enough events should show clusters where potholes likely exist. You would think cities would love that kind of data.

jwoy 8mo ago
This may sound a little cynical but cities often don't like having an inventory of infrastructure problems because it creates an obligation to address it or pay for tire/rim damage if there was prior written notice of a problem. I saw this happen with sidewalk condition data as another example.

djoldman 8mo ago

How happy would HN be if, instead of buying location data as an end run around the 4th amendment, states and municipalities bought traffic data and used it to make safer streets and highways?

killjoywashere 8mo ago

Table 1 and Figure 4: why are the Virginia controlled-access highways different? That population stands out in a way that smells like either a cultural difference or a policy difference. I spend most of my time in California, have lived in SoCal and NorCal, and I spend a fair chunk of time driving around Virginia. My guess is that there's something fishy with the Virginia data being reported. Because if there is anyplace on earth with an insane number of controlled access roads, it's gotta be NVA/DC metro area (or the Tri-Border Area as I like to call it). Also, they need to either update the caption for Figure 4, or move the plots to correspond with the caption. Clearly the Virginia data is on top (or the code is wrong, which seems exceedingly unlikely).

1970-01-01 8mo ago
Speeding in VA just as you would in all other states literally becomes a criminal offense if you're caught. The data is reflective of that.

cabirum 8mo ago

> information from the Android Auto platform Can the quantity and length of tire marks on tarmac be used as a proxy for HBEs?

v4rp1ng 8mo ago

what I thought when I read the article: wait, what? they're collecting this information over my phone from my car!? eventually got me to this: https://arxiv.org/abs/2409.15561 https://arxiv.org/abs/2409.15561 maybe interesting for someone whos privacy focused.

sspehr 8mo ago

Google Maps is definitely one of the technically coolest big tech products imo. It's fascinating what kind of non-obvious insights you can get from the combined user data (E.g. Crowdedness of public places, insights like in the article etc.)

4gotunameagain 8mo ago
The value of said data is tremendous, and could lead to great societal benefits if available for research. So much so that I would argue that google (et al) should be compelled to make the data publicly available. They have enough money, let's make society better.
- sagarm 8mo ago
  They are making this data publicly available -- after aggregation and anonymization.
  - 4gotunameagain 8mo ago
    A fraction of the data they collect, to keep up appearances.
    - notatoad 8mo ago
      Publicly releasing 100% of the maps data they collect would be the most gigantic privacy violation the world has ever seen.
      - 4gotunameagain 8mo ago
        Are you seriously suggesting that google is gatekeeping the data because they care about privacy ?
timmg 8mo ago
What I find very interesting about it is how much data they have (and are able to collect). I'm personally someone who is less worried about "privacy" for this sort of data. But I know lots of other people feel differently. To me, this is a great example of the "greater good" for data sharing. But it is also a great example of responsible use of data.

zx8080 8mo ago

Nice money they can made selling to insurers the collected data on who drives frequently on those road segments with high risk. There's even no need to get the real names! To anonymously load person's risk coefficient, it's enough to ask to log in using Google account when buying insurance.

HungrMn 8mo ago

[dead]

spacecadet 8mo ago

Defensive driving. Two hands, eyes on the "road", not caring about fools and maintaining safe distances. You have all the time in the world... I Recommend taking your car to a HPDE event, sometimes insurance will discount you for this. In this case, hard braking, is you just not knowing anything about your car or the cars around you... I barely use my brakes, but I also drive manual... Id say learn engine braking but y'all probably driving those stupid over powered electric cars...

whynotmaybe 8mo ago
https://donmcminn.com/2020/02/we-judge-ourselves-by-our-intentions-and-others-by-their-behavior/ https://donmcminn.com/2020/02/we-judge-ourselves-by-our-inte...
- spacecadet 8mo ago
  Nice, but... We live in a world with shared roads... I think a much worse outcome is unfolding as we sell self-driving/assisted EVs to people with little to no real driving experience, while also not upgrading the roads, the laws, and driving education.

1970-01-01 8mo ago

My gut tells me there's also another signal in the data. Soft braking events under conditions that call for acceleration. E.g. driver cannot process a highway merge in real time, so they slow down to have more time to think about what they should do next. Go!

thefringthing 8mo ago

> The core objective was to determine if a high frequency of HBEs causally links to a high rate of crashes. We employed negative binomial (NB) regression models [...]. A negative binomial regression fit on observational data cannot establish that a statistical relationship is causal.

bcrl 8mo ago

Many of the merge lanes in California are insanely short compared to those in the rest of the world. The worst are the ones that have merge immediately before an overpass and exit immediately after where merging and exiting have about the width of the overpass to change lanes. I found those infuriating when I used to visit friends in the Bay area. The pattern where I live is the opposite (long exit lane before the overpass and a long merge lane after) and provides far better margins of safety.