The Map Wars · Part 1
Scan the map industry's 2026 headlines: an open-map alliance says it will keep its data fresh with members' "heartbeat signals"; a European mapmaker used an antitrust complaint to break up Google's bundle; HERE's assets are waiting for a home; carmakers buy map data with one hand while plotting to build their own with the other.
Four headlines, four reruns. Wind the clock back forty years and the same scripts played out in full. The lead was a company called Tele Atlas — dissolved fifteen years ago, yet almost everything happening in this industry today has an original version somewhere in its story.
Half a Million Dollars and a Cassette Tape
Start with a sailboat race.
In 1983, SRI researcher Stan Honey navigated Atari founder Nolan Bushnell's boat to victory in the Transpac race. Bushnell, dazzled by Honey's home-built navigation electronics, asked what else he had in mind. Car navigation, Honey said. Bushnell wrote a $500,000 seed check, and the company took its name — Etak — from a Polynesian navigation term.
In 1985 the Etak Navigator shipped: the world's first automotive electronic navigation system. No GPS — dead reckoning, wheel-mounted sensors, a CRT screen on which the car stayed fixed while the map moved. The maps were stored on cassette tapes. It made the cover of Popular Science.
In 1989, Rupert Murdoch's News Corp bought Etak for $25 million. His logic is worth copying down: "where's the nearest X, and how do I get there" would personalize classified ads and yellow pages, making the digital map an advertising medium. That was 1989 — fifteen years before Google turned the map into a local-search gateway. The right vision on the wrong vehicle equals no vision. Insight is cheap; timing is expensive.
Etak changed hands to Sony in 1996. According to the antitrust complaint Tele Atlas later filed against Navteq, Sony under-invested, Etak lost its edge, and Navteq was allowed to run away with the market. Remember that detail — the mapmaker suing the market's dominant player is a family tradition in this industry. It will come up again.
The Godfather of Ghent
On the other side of the Atlantic, a second bloodline was growing.
Tele Atlas was founded in 1984 in 's-Hertogenbosch, the Netherlands. But the company's soul came from Ghent, Belgium: Alain De Taeye — an engineering-architecture graduate of Ghent University and an operations-research academic — had built one of the first digital maps on a PC in the early eighties and founded I&M to make map databases and route planning. I&M merged into Tele Atlas in 1989; from 1990 De Taeye ran the whole group. Ghent's tech scene still calls him its godfather; Tele Atlas was the city's first large technology company.
What followed was textbook expansion: Bosch took a stake (the interests of the in-dash navigation era), listings in Frankfurt and then Amsterdam — the 2005 IPO valued the company at roughly €1.7 billion. In May 2000 it bought Etak from Sony and made it the North American arm; in 2004 it acquired America's GDT with $210 million of institutional backing to complete US coverage. That left exactly two companies on Earth able to make a global map: Tele Atlas strong in Europe, Navteq strong in North America. The duopoly was counted out street by street, and bought out deal by deal.
"Heartbeat Signals," Vintage 2008
Now the first rerun.
On June 30, 2008, Tele Atlas signed a five-year agreement with Google: 200 countries, supplying Google Maps, Google Earth and mobile. Buried in the deal was a clause that looked minor at the time — Tele Atlas would receive back the edits made by Google Maps' user community. Combined with MapShare feedback from twenty million TomTom devices, the October 2008 release of MultiNet became the world's first map database validated by community input: more than 50,000 street-name and traffic-flow corrections verified by the community in a single release — matching the detection volume of all its other 50,000 sources combined.
Devices as probes, community as verification, business exhaust feeding map freshness — sound familiar? The "heartbeat signals" story Overture tells in 2026, Tele Atlas invented in 2008, logic unchanged, word for word. What it lost was never the idea; it was probe count: twenty million PNDs against billions of smartphones, two orders of magnitude apart. Ideas are cheap; distribution is expensive. That is lesson one.
An Exclusive That Lasted Thirteen Months
The second rerun was faster, and it hurt more.
In September 2008, Google Maps dropped Navteq worldwide and switched exclusively to Tele Atlas — the immediate trigger being Nokia's purchase of Navteq; Google would not leave its data lifeline in a rival's hands. The mapmaker had landed the biggest map customer on Earth.
Thirteen months later, in October 2009, Google announced it was dropping Tele Atlas in the United States in favor of its own data: road networks from the Street View fleet, public census data, and crowdsourced fixes from a "Report a Problem" button. Barely a year after signing a long-term agreement. Country-by-country replacement followed until the relationship was essentially wound down.
The scene deserves to be carved above every data company's door: selling your data to a platform customer capable of building its own is tutoring your replacement. Google used Tele Atlas's map as a crutch while it learned to walk, then threw the crutch away.
The symmetrical footnote came three years later: when Apple Maps launched in 2012, its base map ran on TomTom — Tele Atlas-lineage — data. Lose Google, gain Apple. Between giants, a mapmaker is always betting on the same question: am I a long-term partner, or a transitional supplier? And the 2026 echo is already audible: Mercedes pipes Google's data into its own navigation; Rivian's navigation is built on Google's SDK. Forty years on, teacher and student have swapped places, but the game hasn't changed a word.
What €2.9 Billion Bought, and What It Left
The 2007 bidding war needs only a few lines: TomTom bid €2 billion in July; Nokia's $8.1 billion purchase of Navteq on October 1 detonated the arms race; Garmin counter-bid $3.3 billion; TomTom raised to €2.9 billion; on November 16 Garmin withdrew and signed a supply deal with Navteq instead — and the loser's stock rose 12.5% that day, more than the winner's. The market voted on the spot.
The year it was bought for €2.9 billion, Tele Atlas's entire estate was: €308 million in revenue, €24.4 million in net profit, 1,889 employees. What followed were three goodwill impairments totaling roughly €1.73 billion, and a debt load that nearly crushed the buyer.
But watch where the asset went: the database called MultiNet did not die. It evolved, generation by generation, into the Orbis layers that sit on Overture's open base map today — the fourth generation of the same asset. Parts of Apple Maps' foundation, the engine behind Microsoft Azure Maps, the navigation running in the cockpits of Chinese cars exported worldwide — trace them back far enough and you reach the streets a group of people in Ghent and 's-Hertogenbosch started counting forty years ago.
Names Die; Assets Move House
Now spread out the family tree.
Etak, builder of the first car navigation device, ceased to exist not long after its acquisition. Tele Atlas was dissolved as an entity in 2011, absorbed into TomTom. Navteq's name vanished too after Nokia swallowed it; the assets were rebadged HERE. Three names, three tombstones, zero grams of asset lost — each time, the name died and the database moved house, feeding the next generation of products.
The human arc closed the same way: De Taeye joined TomTom's management board after the acquisition and ran its maps business until April 2026, exiting at the same shareholder meeting as TomTom's founder. From a PC in 1980s Ghent to a 2026 AGM — one man walked the industry's entire cycle.
Which is why the judgment made elsewhere in this series about HERE — it won't go bankrupt, but it will disappear — is no bold prophecy. It is merely the third execution of this industry's fixed sentence structure: Etak was the first, Tele Atlas the second. In mapping, the lifespan of a name and the lifespan of a company have never been the same thing.
Forty Years, One Unchanged Problem
The industry's fundamental business problem has not moved since 1984: collection costs are enormous, marginal reproduction costs are zero, and your biggest customer can become your biggest rival at any moment.
Four generations turned in four answers. Etak sold itself to a media mogul and bet on advertising — right vision, wrong vehicle. Tele Atlas sold itself to a device maker and bet on vertical integration — just in time for the category's extinction. Navteq sold itself to a phone maker, which passed it to carmakers, betting on HD maps — and the narrative receded. The fourth answer is the one TomTom is writing now: donate the base map to the commons and bet that value migrates up to the responsibility and dynamic layers. Verdict unknown — but it is the first answer that doesn't involve selling yourself.
There are no new stories in the map industry. Heartbeat signals are 2008's community validation; platform self-build is Google's 2009 graduation; the antitrust suit is a sequel to Tele Atlas v. Navteq, 2005; HERE's destination is a road Etak and Tele Atlas already walked. But no new stories does not mean no new endings — the first three cycles all closed with a name disappearing, and the protagonist of the fourth cycle has, for the first time, chosen not to sell. Whether that ending is worth watching, 2027 will tell.
Sources: Etak oral history (IEEE Engineering & Technology History Wiki); Smithsonian National Museum of American History collection records; Tele Atlas v. Navteq court filings (N.D. Cal.); TomTom/Tele Atlas press archives; Search Engine Land and Blumenthals coverage of Google Maps' 2008–2009 data transitions; Geospatial World profile of Alain De Taeye.
《地图战争》系列 · 第1篇
看一眼2026年地图行业的新闻清单：开放地图联盟宣布用成员的"心跳信号"维持数据鲜度；欧洲图商靠反垄断投诉拆掉了谷歌的捆绑；HERE的资产在等一个归宿；车厂一边买图商的数据、一边琢磨自建。
四条新闻，每一条都是重播。把时间拨回四十年，同样的剧本完整地演过一遍。主角叫Tele Atlas——这家公司已经注销十五年了，但今天行业里发生的每件事，几乎都能在它的故事里找到原版。
五十万美元和一盘磁带
先从一场帆船赛讲起。
1983年，SRI研究员Stan Honey在跨太平洋帆船赛上给Atari创始人Nolan Bushnell当领航员，夺冠。Bushnell被他自制的导航电子设备折服，问他还有什么想法。Honey说，想做汽车导航。Bushnell给了50万美元种子钱，公司取名Etak——波利尼西亚的航海术语。
1985年，Etak Navigator问世，全球第一台车载电子导航：没有GPS，靠航位推算，轮内磁条传感器计程，CRT屏幕上车不动、地图动——而地图，存在磁带上。上市那年登了《大众科学》封面。
1989年，默多克的新闻集团花2500万美元买下Etak。默多克的逻辑值得抄下来：他认为"哪里最近、怎么去"会让分类广告和黄页变得个人化，数字地图会成为重要的广告媒介。这是1989年——比谷歌把地图变成本地搜索入口早了十五年。正确的愿景配错误的载体，等于没有愿景；洞察不值钱，时机才值钱。
1996年Etak转手索尼。按Tele Atlas后来起诉Navteq垄断的诉状说法，索尼投入不足让Etak失去竞争力，放任Navteq坐大。记住这个细节——图商起诉市场支配者，是这个行业的传统艺能，后面还会出现。
根特的教父
大西洋另一边，另一条线在生长。
1984年，Tele Atlas在荷兰斯海尔托亨博斯创立。但这家公司的灵魂来自比利时根特：Alain De Taeye，根特大学工程建筑专业出身、商学院运筹学研究员，八十年代就在一台PC上做出了第一张数字地图，随后创办I&M公司做地图数据库和路径规划。1989年I&M并入Tele Atlas，1990年起他执掌整个集团——根特科技圈至今叫他"教父"，Tele Atlas是这座城市第一家大型科技公司。
之后是教科书式的扩张：博世入股（车载导航时代的利益绑定），先后登陆法兰克福和阿姆斯特丹交易所，2005年上市时估值约17亿欧元。2000年5月从索尼手里买下Etak改组为北美分部，2004年再收美国GDT、机构注资2.1亿美元补齐美国覆盖。至此全球只剩两家能做全球地图：Tele Atlas强欧洲，Navteq强北美。双寡头格局，就是这么一条街一条街数出来、一家一家买出来的。
2008年就有的"心跳信号"
现在讲第一次重播。
2008年6月30日，Tele Atlas和谷歌签下五年协议：覆盖200个国家，供给谷歌地图、谷歌地球和移动端。协议里有一条当时看着不起眼的条款——Tele Atlas可以拿回谷歌地图用户社区的编辑数据。加上母公司TomTom两千万台设备的MapShare反馈，2008年10月发布的MultiNet成为全球第一个用社区数据验证变更的地图数据库：单个版本里，靠社区输入验证的街名和通行方向修正超过5万处，检出量追平了其他全部五万个数据源的总和。
终端当探针，社区当校验，业务副产品反哺地图鲜度——是不是很耳熟？Overture在2026年讲的"心跳信号"故事，Tele Atlas在2008年就发明了，连逻辑都没改过一个字。它输掉的从来不是理念，是终端规模：两千万台PND对几十亿部智能手机，探针数量差了两个数量级。理念不值钱，分发才值钱，这是第一课。
十三个月的独家
第二次重播更快，也更疼。
2008年9月，谷歌地图全球弃用Navteq、独家切换到Tele Atlas——直接动因是诺基亚买了Navteq，谷歌不肯把数据命脉放在对手手里。图商拿下了世界上最大的地图客户。
十三个月后，2009年10月，谷歌宣布在美国停用Tele Atlas，改用自有数据：Street View车队采的路网、人口普查的公开数据、加上"报告问题"按钮收上来的众包修正。距离双方签下长期协议，一年零三个月。之后逐国替换，直至基本清退。
这一幕值得刻在每家数据公司门口：把数据卖给一个有能力自建的平台客户，等于给自己的替代品当老师。谷歌用Tele Atlas的图当拐杖学会了走路，然后把拐杖扔了。
对称的注脚在三年后：2012年苹果地图上线，底图用的正是TomTom——Tele Atlas血统的数据。失谷歌、得苹果，图商在巨头之间赌的永远是同一件事：自己是长期伙伴，还是过渡供应商。而2026年的回声已经响起来了：奔驰把谷歌的数据接进自家导航，Rivian的导航建在谷歌的SDK上——四十年后，师生关系换了方向，游戏一个字没变。
29亿欧元买走的和留下的
2007年的竞购战只需要几行：7月TomTom出价20亿欧元；10月诺基亚81亿美元买Navteq引爆军备竞赛；Garmin抬价到33亿美元；TomTom反手加到29亿欧元；11月16日Garmin退出、转签Navteq保供协议——当天退出者股价涨了12.5%，比赢家涨得多。市场当场就投了票。
被29亿欧元买下那年，Tele Atlas的全部家底是：3.08亿欧元营收，2440万净利润，1889名员工。之后的故事是三笔合计约17.3亿欧元的商誉减值，和一段差点压垮买家的债务。
但注意资产的去向：那套叫MultiNet的数据库没有死。它一路演进，到今天叠在Overture开放底图上的Orbis，是同一套资产的第四代。苹果地图的部分底层、微软Azure Maps的引擎、中国车企出海车型座舱里跑的导航——追根溯源，都是根特和斯海尔托亨博斯那批人四十年前开始数的街。
名字死了，资产搬家
现在把家谱摊开看。
Etak，1985年造出第一台车载导航，被收购后"不久即不复存在"。Tele Atlas，2011年作为实体注销，品牌吸收进TomTom。Navteq，被诺基亚吞下后名字也消失了，资产改叫HERE。三个名字，三块墓碑，零克资产损失——每一次，名字死了，数据库搬了家，继续给下一代产品供血。
人的弧线同样闭合：De Taeye在收购后进入TomTom管理委员会执掌地图业务，一直干到2026年4月，与TomTom创始人Goddijn同一场股东大会退出。从八十年代根特的一台PC，到2026年的谢幕，一个人走完了这个行业的整个轮回。
所以此前我们对HERE的那个判断——不会倒闭，但会消失——根本算不上大胆预言。它只是这个行业固定句式的第三次执行：Etak是第一次，Tele Atlas是第二次。名字的寿命和资产的寿命，在地图行业从来是两回事。
四十年，一道没变的题
这个行业的根本商业问题，从1984年到2026年没变过：采集成本巨大，边际复制成本为零，而最大的客户随时可能变成最大的对手。
四代人交了四份答卷。Etak把自己卖给了媒体大亨，赌广告——载体错了。Tele Atlas把自己卖给了设备商，赌垂直整合——赶上了品类灭绝。Navteq把自己卖给了手机厂再转给车厂，赌高精地图——叙事退潮了。第四份答卷是TomTom正在写的：把底图捐给公共品，赌价值上移到责任层和动态层——对错未知，但它至少是第一份不靠"把自己卖掉"来作答的答卷。
地图行业没有新故事。心跳信号是2008年的社区验证，平台自建是2009年的谷歌毕业，反垄断诉讼是2005年Tele Atlas诉Navteq的续集，HERE的归宿是Etak和Tele Atlas走过的路。但没有新故事不等于没有新结局——前三个轮回都以名字消失收尾，第四个轮回的主角第一次选择了不卖。这个结局值不值得看，2027年见分晓。
数据来源：Etak口述史（IEEE工程技术史维基）、史密森尼美国历史博物馆藏品档案、Tele Atlas诉Navteq案卷（N.D. Cal.）、TomTom/Tele Atlas历年新闻稿、Search Engine Land与Blumenthals对2008-2009年谷歌地图数据切换的报道、Geospatial World对De Taeye的人物档案。