雪山腳下
走馬觀花
外媒報道，美國亞利桑那州大學的人機界麵與控製實驗室正在開發一種腦力控製的無人機導航界麵，用腦電波來指揮一群無人機工作。
目前的無人機主要通過操縱杆或者手機操控，這意味著一位飛行員隻能夠同時操控一架無人機。借助腦機接口技術，一位飛行員就能夠同時操控多架無人機，讓它們形成一個編隊，或者讓它們分散在不同的飛行路線上。
根據美國國防部先進研究計劃署（DARPA）提供的項目摘要，該研究的最終目的是找到人類大腦“感知多代理係統信息”的運行機製，繼而從中提取“控製命令”。換句話說，科學家們正在試圖建立一個係統，讓一個人可以控製一群無人機，並且這些無人機還能接收不同的命令，做出不同的行為。
該項目的負責人Panagiotis Artemiadis介紹，該研究將分為幾個步驟進行：“首先我們對人類進行訓練，以便他們能夠想象出這些行為。然後我們訓練出一種算法，將這些行為與大腦不同部分的激活聯係起來。”
Artemiadis認為，開發這套腦力控製係統有著很現實的優勢，它可以讓一個操作員同時控製多個看似獨立的無人機執行不同的任務。大腦操控的無人機能夠提升搜索和救援任務，能夠比單一無人機更有效的搜索更大區域範圍。
譬如，利用無人機同時檢查100台風力渦輪機，或者更高效地進行搜索和救援任務；無人機機群或許也能夠用於幫助人們撲滅野火，它們能夠追蹤大火的速度，並且為現場急救員和消防員提供大量的照片和數據，這是目前的技術無法做到的；在大型集會的現場，無人機機群能夠進行實時監控，並且向監管部門發送更廣區域範圍內的實時視頻記錄。
Artemiadis稱，這種“集群控製”的行動模式是從鳥類和魚類等自然群體的行為中獲得的靈感。“我們可以提取與所期望的集體行為相關的信息，比如以特定的編隊飛行，這在普通的人工控製界麵上是不可能實現的。”
據介紹，按照上述研究，無人機與人將通過腦電波的傳送來實現通訊。
“隻要有一個集中控製器可以將命令傳達給無人機，腦群控製算法甚至可以擴展到同時控製成百上千的無人機。” Artemiadis說道，目前他已經成功完成了同時控製3台無人機的測試。
Artemiadis的無人機研究受到美國國防部先進研究計劃署和美國空軍科學研究辦公室的資金支持。美國國防部先進研究計劃署認為，腦力控製的無人機編隊是相當安全的，它的實現隻是個時間問題。
原文鏈接：https://www.sciencedaily.com/releases/2017/09/170929230433.htm
Brain-controlled drones are here: What’s coming in the next five years?
-
Date:
-
September 29, 2017
-
Source:
-
Arizona State University (ASU)
-
Summary:
-
Single unmanned autonomous vehicles (UAVs) directed by joysticks, radio controllers, and mobile phones are already accomplishing a variety of useful tasks, such as aerial photography and security patrols. But using multiple drones requires multiple human operators, and this presents a coordination problem.
FULL STORY
Single unmanned autonomous vehicles (UAVs) directed by joysticks, radio controllers, and mobile phones are already accomplishing a variety of useful tasks, such as aerial photography and security patrols. But using multiple drones requires multiple human operators, and this presents a coordination problem.
Now a single operator using emerging human-brain interfaces can control a swarm of drones, making possible new classes of applications, according to Panos Artemiadis, director of the Human-Oriented Robotics and Control (HORC) Lab at Arizona State University.
Artemiadis thinks it is likely that drone swarms using human-brain interface mechanisms will, in the next three to five years, make inroads where individually controlled UAVs cannot. Here are a few of the drone applications that are now within reach:
Search and Rescue Missions
Humans will collaborate with swarms of robots in search and rescue scenarios. The brain-robot interface enables control of many robots at the same time, and it scales the ability of a robotic team to cover larger areas in less time. If the controller detects something in the video stream that warrants closer surveillance, the swarm can be directed to close in on that area.
Fire Fighting
Armed with infrared imaging equipment, a drone swarm can be used to track the spread of a forest fire over large areas in real time, allowing firefighters to adjust their plans accordingly. The human controller can follow a reported change in weather conditions, such as a shift in wind direction, with a swarm of drones to determine if the fire has jumped to a new area.
Agriculture Analysis
Teams of drones will oversee and analyze large agricultural fields -- creating topographic maps for soil analysis and irrigation planning. In addition to being outfitted with cameras, aerial drones will use sensors to identify necessary irrigation adjustments and scanners that can identify crop infections or infestations. Some drone systems are already being used for crop spraying -- swarms will be able to accomplish the task more quickly and efficiently.
Entertainment
As drones enter the entertainment arena, we will begin to see mind-controlled drone swarms for events. For example, a single person could operate a fleet of drones shooting photos and videos at an outdoor concert or sports venue, narrowing in on spectator activities for display on the Jumbotron. And while Lady Gaga's Super Bowl drones were controlled by a central computer (and filmed in advance of the show), smaller swarms can be managed by a single human for smaller light displays or to drop gifts (t-shirts or CDs, for example) into a crowd.
Cyber-physical surveillance systems
Understanding brain-drone interfaces allows building cyber-physical surveillance systems that combine human intuition and experience with the sensing capabilities of multiple drones. This would allow more efficient and accurate surveillance systems than what is now available -- especially for large, security sensitive events like bowl games, marathons and political rallies.
Story Source:
Materials
provided by Arizona State University (ASU). Note: Content may be edited for style and length.
轉自微信公眾號兒"中外學術情報"
