•  READ MORE: Could a 核兵器 save us from a 殺し屋 asteroid?

|

天文学者s and data scientists working with a 上げる in computer 力／強力にする on 貸付金 from Google have discovered 27,500 new asteroids, some perilously の近くに to Earth.

Their 共同 hopes to 速度(を上げる） up the 開発 of 'a 包括的な 地図／計画する of the solar system' needed for 'planetary 弁護,' によれば one Harvard astrophysicist, Matthew Holman, who helped develop the asteroid-追跡（する）ing ソフトウェア.

Nearly 1.7 billion points of light, 文書d in 412,000 infrared images from the US 国家の 光学の-Infrared Astronomy 研究 研究室／実験室 (NOIRLab) 古記録s, were scanned by the 事業／計画（する）'s novel '殺し屋 asteroid'-追跡（する）ing algorithm.

'This is 最高の important,' as one former 航空宇宙局 宇宙飛行士 主要な the 事業／計画（する） put it.

'This is the 重要な to 保護するing the Earth from 存在 攻撃する，衝突する by asteroids: knowing where all these are.' 

Since 2002, that retired 航空宇宙局 宇宙飛行士, Stanford-educated astrophysicist Dr. Ed Lu, has served as the (n)役員／(a)執行力のある director of the nonprofit B612, where he now leads the group's Asteroid 学校／設ける.

'At this moment, we are now one of the largest discoverers of asteroids on Earth,' Dr. Lu said, discussing the team's 身元確認，身分証明 of tens of thousands of new asteroids.

'But what makes this 利益／興味ing is that we don't own a telescope. We don't operate a telescope,' he continued. 'We're doing this from a data science 視野.'

The B612 Asteroid 学校／設ける's algorithm ― known as the Tracklet-いっそう少なく Heliocentric 軌道 回復, or THOR ― trawled NOIRLab's astronomy image 古記録 to distinguish which 反映するd points of light were really nearby asteroids.

THOR could calculate, based on 証拠 from just two images, いつかs taken on different nights and 経由で two different telescopes, if those points of light were, in fact, one asteroid in 軌道 around our solar system.  

'What is exciting is that we are using 電子s in data 中心s, in 新規加入 to the usual photons in telescopes, to make 天文学の 発見s,' によれば Dr. Lu.

The THOR algorithm was （手先の）技術d by B612's Asteroid 学校／設ける in 共同 with the University of Washington's DiRAC 学校／設ける.

The raw 計算するing 力／強力にする needed to identify viable asteroid 候補者s across these disparate telescopic images would have once made the 過程 all but impossible, even in the 最近の past.  

But Google Cloud's 分配するd 網状組織 of 計算するing 力／強力にする 許すd THOR to identify its 27,500 new, 'high-信用／信任 asteroid 発見 候補者s' in only about five weeks.

'This is an example of what is possible,' Massimo Mascaro, a technical director for Google Cloud's office of the 長，指導者 科学（工学）技術 officer, said.

'I can't even quantify how much 適切な時期 there is ーに関して／ーの点でs of data that is already there collected, and, if 分析するd with the proper computation, could lead to even more results,' Mascaro explained.

によれば a 圧力（をかける） 声明 from the B612 nonprofit, their 共同 with Google Cloud will 結局 put THOR to work on 5.4 billion 観察s from different 天文学の 調査するs and 器具／備品 ― celestial mysteries which may 証明する to be asteroids or other outer space 反対するs.

The Asteroid 学校／設ける 発表するd that it is also 調査するing the 実現可能 of 雇うing Google's AI 科学（工学）技術s in an 成果／努力 to 自動化する the vetting and 立証 of 可能性のある asteroid 候補者s, as first 報告（する）／憶測d by their THOR algorithm. 

In the past, the 初期の 立証 of these 候補者s has been left to a volunteer army of high school students, under 卒業生（する） college students, postgraduate 研究員s, scientists, and professional 天文学者s. 

But, if these AI-補助装置d 拡大s of their 事業／計画（する） are successful, Dr. Lu said that the 削減 in this time-消費するing human 立証 work will help the 学校／設ける adapt THOR's 過程 for much more 広大な and expansive astronomy datasets, like those coming soon from the Vera C. Rubin 観測所, which now under construction in Chile.

'This is a sea change,' Dr. Lu said.

Asteroid hunters 位置／汚点／見つけ出す more than 27,000 'hidden' space 激しく揺するs - and some pass '危険に' の近くに to Earth