Go homepage(回首页) Upload pictures (上传图片) Write articles (发文字帖)
The author:(作者)delvpublished in(发表于) 2013/12/29 5:13:15 Former engineer reality： Google maps calculate navigation routes
Former engineer reality: calculate navigation routes-Google maps, Google maps, Google Maps, Google-IT information Former engineer reality: Google maps calculate navigation routes
A former Google engineer lichade·lasaier (Richard Russell) represented on the social questions and answers site Quora, when users search for routes using Google maps, user factors include various aspects of the final route.
Russell said: " with other products similar to Google maps route decisions are also based on a number of factors, and depend on the data that is available in a particular area. These factors include the official maximum speed limit and suggested driving speeds, road types a given speed, the historical average speeds within a certain period of time (sometimes taking the average, sometimes the value is within a particular time of the day), the user's actual travel time, as well as real-time traffic information. Using that data from different sources and make better predictions . ”
Most companies that offer real-time traffic forecasts with the actual results will be compared so as to continuously improve the algorithm and data sources. The company will do better in this regard can be predicted in the medium and long term to provide the best features.
In short, Google's calculation of a large amount of data, then tell the user destination is 10 minutes by car . In 2009, Google had disclosed crowdsourced traffic data how to make Google maps more accurate.
Google maps product manager daifu·BASI (Dave Barth), said at the time: "when you compare speed and the speed of other phones on the road, and at any time to obtain information of thousands of mobile phones in the city, we have a good understanding of real-time traffic. ”
Of course, regardless of the data introduced, drive results that were not perfect. Russell says calculating route "is a problem of predicting the future, and follow certain patterns in traffic, but in the nature is unpredictable." Google could understand large amounts of information, but had no way of knowing if a traffic incident occurs, or is there a school bus in front of the user-driven car traffic.
前工程师揭秘：谷歌地图如此计算导航路线 - 谷歌地图,Google Maps,谷歌 - IT资讯前工程师揭秘：谷歌地图如此计算导航路线
一名前谷歌工程师理查德·拉塞尔（Richard Russell）在社交问答网站Quora上表示，当用户使用谷歌地图搜索路线时，决定用户最终路线的因素包括多个方面。
拉塞尔表示：“与其他产品类似，谷歌地图的路线决策也基于一系列因素，并取决于某一特定地区的可用数据。这些因素包括官方最高限速和建议驾驶速度、根据道路类型推算出的速度、在某段时间内的历史平均速度（有时取平均值，有时是一天内某段特定时间的值）、此前用户的实际行驶时间，以及实时路况信息。他们将这些不同来源的数据混合，并做出最好的预测。”
大部分提供实时路况的公司都会将预测与实际结果进行对比，从而不断改进算法和数据源。在这一方面做得较好的公司将可以在中长期内提供最好的预测功能。
简而言之，谷歌计算了大量数据，随后才能告诉用户目的地车程为10分钟。2009年时，谷歌还曾披露众包的交通流量数据如何使谷歌地图更准确。
谷歌地图产品经理戴夫·巴斯(Dave Barth)当时表示：“当我们将你的速度与道路上其他手机的速度进行对比，并在任意时刻都获得城市中数千部手机的信息时，我们可以很好地了解实时路况。”
当然，无论引入什么样的数据，得出的车程结果可能都不是完美的。拉塞尔表示，计算路线“是一个预测未来的问题，而交通流量尽管遵循一定模式，但从本质来看是不可预测的”。谷歌可能了解大量信息，但不可能知道是否有交通事故发生，或者是否有校车在用户驾驶的汽车前缓慢行驶。
赞