2021年经济学人 如何统计全球太阳能装置数目(下)

（单词翻译）

That is where the second technological¹ trend comes in.

这就体现了第二种技术趋势的用武之地。

Dr Kruitwagen and his colleagues trained a machine-learning system to spot the solar panels for them.

克鲁特瓦根博士和他的同事开发了一个机器学习系统来为他们识别太阳能电池板。

Computer vision is a hot field.

计算机视觉是一个热门领域。

But the specifics of orbital reconnaissance meant that off-the-shelf software was not suitable for the task the researchers had in mind.

但轨道侦察的特殊性意味着现成的软件不一定能完成研究人员心目中的任务。

Machine-learning systems are taught what to do by examining a "training set", which contains examples of what is being searched for.

机器学习系统通过检查“训练集”来学习该做什么，“训练集”包含正在搜索的内容的示例。

For common tasks such as facial recognition, pre-built training sets are often available.

对于面部识别等常见任务，通常可以使用预先构建的训练集。

But Dr Kruitwagen's team had to build their own.

但是克鲁特瓦根博士的团队必须建立他们自己的“训练集”。

For this, they turned to OpenStreetMap, an open-source rival to Google Maps in which volunteers had already tagged large numbers of solar plants.

为此，他们求助于OpenStreetMap——谷歌地图的开源竞争对手，志愿者已经在这个平台上标记了大量的太阳能发电厂。

But there was little consistency².

但几乎没什么一致性。

"Some people had just drawn³ rough outlines around an entire field,"Dr Kruitwagen says.

克鲁特瓦根博士说：“有些人只是粗略地勾勒出了整个领域的轮廓。”

"Others had gone in and traced the outline of each row of panels separately."

“其他人走进去，分别画出了每排嵌板的轮廓。”

Fixing that involved a great deal of manual labour.

要解决这个问题，需要大量的体力劳动。

Once the training data had been cleaned up, the learning algorithms had to be tweaked as well.

一旦训练数据被清理干净，学习算法也必须进行调整。

From space, even big solar installations look small.

从太空上看，即使是大型的太阳能装置也看起来很小。

Each pixel in the Sentinel images represented a ten-by-ten-metre square.

“哨兵”号图像中的每个像素代表一个10*10米的正方形。

Even for the higher-resolution SPOT satellites, the squares’ sides are one and a half metres long.

即使是高分辨率的SPOT卫星，正方形的边长也有1.5米。

Existing classifiers, trained for things like facial recognition or self-driving cars, are used to spotting objects that loom⁴ large in their field of vision.

现有的分类器受过面部识别或自动驾驶汽车等方面的训练，可用于发现在他们的视野中逼近的物体。

Hunting for smaller ones meant tinkering with the software to boost its ability to detect tiny features.

寻找较小物体意味着对软件进行升级，以提高其检测微小特征的能力。

False positives—things like tennis courts and agricultural greenhouses that resemble solar panels from space—had to be removed.

误报信息——比如网球场和农业大棚，从太空看起来像太阳能电池板——必须被移除。

Though extraordinary, Dr Kruitwagen’s results are already out of date.

尽管克鲁特瓦根博士的研究成果非同寻常，但他已经过时了。

The data-gathering phase of the project ended in 2018, meaning that the thousands of new plants built since then are not included.

该项目的数据收集阶段于2018年结束，这意味着自那以后建造的数千座新装置不包括在内。

But the project, he says, proves that the method works.

但他说，这个项目证明了这种方法是有效的。

He intends to make his results, including the labour-intensive training set, available for others to use.

他打算将他的成果，包括劳动密集的“训练集”，提供给其他人使用。

One logical extension of his project, he says, would be to expand the analysis to include solar panels installed on domestic rooftops.

他说，他的项目的一个合理延伸是扩大分析范围，包括安装在家庭屋顶上的太阳能电池板。

Such “behind-the-meter” installations are particularly tricky⁵ to track in other ways.

这样的“表后”装置在其他方面尤其难以追踪。

More generally, Dr Kruitwagen hopes that his eye-in-the-sky approach—which, despite the planetary scale of the project, cost only around $15,000 in cloud-computing time—could presage⁶ more accurate estimates of other bits of climate-related infrastructure⁷, such as fossil-fuel power stations, cement plants and terminals for ships carrying liquefied natural gas.

更广泛地说，克鲁特瓦根博士希望他的“天空之眼”方法——尽管该项目是规模庞大，但在云计算时间上的花费仅为15000美元——能够更准确地估计其他与气候相关的基础设施，如化石燃料发电站，水泥厂和运输液化天然气的船舶码头。

The eventual⁸ result could be the assembly of a publicly available, computer-generated inventory⁹ of every significant bit of energy infrastructure on Earth.

最终的结果可能是列出一份公开可用的、由计算机生成的地球上每一个重要能源基础设施的清单。

Quite apart from such a model's commercial and academic value, he says, an informed public would be one better able to hold politicians’ feet to the fire.

他说，撇开这种模式的商业和学术价值不说，一个知情的公众更能让政客们按捺不住。