Robot dogs could patrol the US-Mexico border

The photos look like a scene out of science fiction: Robot dogs patrolling the US-Mexico border, climbing over harsh terrain to search for threats and contraband.

But these images are real.
The Department of Homeland Security recently released them as it revealed details about how it’s testing the technology.
Officials praised the robots’ potential as a “force multiplier” that could boost Border Patrol agents’ safety by reducing their exposure to life-threatening hazards. An article touting the tests on the DHS Science and Technology Directorate’s website notes that someday the dogs, officially known as Automated Ground Surveillance Vehicles, could become “a CBP agent or officer’s best friend.”
“Don’t be surprised,” it says, “if in the future we see robot ‘Fido’ out in the field, walking side-by-side with CBP personnel.”
But the details about the testing did seem to catch some people by surprise, sparking a flurry of reactions on social media comparing the images to dystopian scenes from sci-fi shows like “Black Mirror.”

Robot vacuum cleaner escapes from Cambridge Travelodge

A robot vacuum cleaner made a break for freedom after giving staff the slip at a Travelodge hotel.

The automated cleaner failed to stop at the front door of the hotel in Orchard Park in Cambridge on Thursday, and was still on the loose the following day. Staff said it just kept going and “could be anywhere” while well-wishers on social media hoped the vacuum enjoyed its travels, as “it has no natural predators” in the wild.

It was found under a hedge on Friday.

Staff at the hotel posted the story of the robot vacuum’s great escape on social media, asking for it to be returned, if found. “Today we had one of our new robot vacuums run for its life,” the assistant manager wrote. “They normally sense the lip at the entrance [to the hotel] and turn around, but this one decided to make a run for it.”

Don’t rush to machine learning

A simpler approach—good data, SQL queries, if/then statements—often gets the job done.

It turns out the best way to do machine learning (ML) is sometimes to not do any machine learning at all. In fact, according to Amazon Applied Scientist Eugene Yan, “The first rule of machine learning [is to] start without machine learning.”

What?

Yes, it’s cool to trot out ML models painstakingly crafted over months of arduous effort. It’s also not necessarily the most effective approach. Not when there are simpler, more accessible methods.

It may be an oversimplification to say, as data scientist Noah Lorang did years ago, that “data scientists mostly just do arithmetic.” But he’s not far off, and certainly he and Yan are correct that however much we may want to complicate the process of putting data to work, much of the time it’s better to start small.

AI活用で不審者検知を、国交省 小田急事件受け

国土交通省は24日、小田急線電車内で起きた乗客刺傷事件を受けてJRや大手私鉄各社などと意見交換し、AIを含む最新技術を活用した不審者検知機能の高度化などの対策をまとめたと発表した。

警備員の巡回など駅や車内の警戒を強化し、ポスターやアナウンスで周知。防犯カメラを増設し、画像解析などで不審者や不審物を検知する技術を関係者間で共有する方策なども検討する。ピクトグラムを使った非常通報装置の使用方法の分かりやすい表示にも努める。

防犯カメラの検知機能を巡っては、JR東日本が同社への重大犯罪で服役した出所者を顔認証機能の検知対象とする方針を撤回したばかり。

少ないデータでAIが作れる技術「スパースモデリング」とは?

「スパースモデリング」と呼ばれるデータ分析手法をご存知だろうか。わずかなデータからでもAIモデルを構築可能な技術であり、2019年にイベント・ホライズン・テレスコープが公開した、ブラックホールシャドウの撮影に利用された手法としても知られる。

スパースモデリングは膨大な量のデータから学習するディープラーニングとは反対に、わずかなデータ量からでもAIを構築可能であり、AIが結論を導く過程が人間にも理解しやすく、AIのブラックボックス化問題の回避も可能だという。

独自のスパースモデリング技術をAIに応用してデジタルソリューションを開発する、AIスタートアップ HACARUSの代表取締役 CEO 藤原健真氏に、同技術の概要と今後の発展性について話を聞いた。