Re: Lahingrobotid, droonid jms distantsjuhitav tehnika
Postitatud: 04 Juul, 2016 17:15
Ilmselt käib ka Delfi siin militaarnetis piilumas, igatahes viivitusega sama teema üleval -> AI toimetab. Delfi muidugi oma tuntud kvaliteedis, proovigu vähemalt terminites rajal püsida.Gideonic kirjutas:Huvitav artikkel AI kohta nimega ALPHA, mis ka peavoo uudistes läbi liikus, et tegi korduvalt ära ka kogenud inimpiloodile:
Kokkuvõte, artikkel ise
Uudistest võis jääda ekslik mulje, et tegu on mingi 1 vs 1 "dogfight" simulaatoriga, inimene vs AI. Tegelikult on ALPHA veel valmimisjärgus olev projekt, mis saab olema pilootide treeningprogramm standardsetes simulaatorites. Sellistes, kus kasutatakse lihtsustamata tõepäraseid (ja ka salastatud) mudeleid lennukite, radarite, sensorite jms töö simuleerimiseks, kusjuures AI saab sensorite info kätte, nagu inimenegi, puudulikult ja müraga ning peab grupis koostööd tegema (igal lennukil oma).
Antud artikli raames keskenduti põhiliselt ühele 2 vs 4 stsenaariumile, kus pahadel (ALPHA) oli kahekordne ülekaal, kuid see-eest väiksema laskeulatusega keskmaa raketid ja puudus AWACS tugi (selged paralleelid Hiina vs USA). Selles stsenaariumis on võimalik ALPHA-l võita ainult vastase üle manööverdades (võimalusel kasutades aruvlist ülekaalu) vastasel juhul lastaks nad kaugemalt lihtsalt alla.
Esialgu võisteldi lihtsama (simulaatori standardse) AI vastu ning iseenda (ALPHA varasemate) versioonide vastu. Kui sellega saavutati teatav tase, panid ühe Blue Force hävitaja tooli arvestava kogemusega endine USAF piloot, Kolonel Gene “Geno” Lee.
Kui simulaatori tavalist AI'd suutis ta samas stsenaariumis probleemideta võita, siis antud ALPHA iteratsioonile jäi ta alla iga kord (Geno lasti alla iga kord, ALPHA juhitud lennukeid ta tabada ei suutnud).
Artiklis on mainitud, et see AI pole veel kaugeltki valmis ning arenguruumi on. Lähitulevikus on plaanis formaalsemad testid inimeste vastu keerulistemates stsenaariumites.
Rõhutan veelkord, et selle AI eesmärk on siiski treening, mitte inimeste väljavahetamine kokpitis. Kui see suudab aga viletsama situatsiooni-teadlikusega (pole AWACS tuge, sitem radar, jne) korduvalt tuhandete lennutundidega inimpilooti alla lasta, saab see valmis kujul olema väga tõsiseltvõetav treeningvahend. Kui värsked 0-lennutundidega juntsud hakkavad pidevalt sellise asja vastu sõdima, võib see olla palju suurem force multiplier, kui misiganes uus tehniline vidin lennukite peal.
Lõpetuseks tasub mainimist, et ALPHA ise jooksis ühe 35$ Rasberry Pi peal, nii et ei vaja tööks mingit superkompuutrit. Soovitan lugeda ise algallikast:
ALPHA’s [the AI's] current primary objective is to serve as an intelligent hostile force for pilots to train against within the AFSIM simulation environment [8]. While the models showcased within this document for platforms do not mimic any system specifically, AFSIM realistically represents a modern air combat environment with appropriately behaving models for aircraft, sensors, and weapons. Data is collected from sensors, fused real-time, and sent to ALPHA, complete with noise and potential failures. This data feed is received at and commands sent to unique UDP ports for each platform, causing this problem to contain some of the software complexities that would be present in actual hardware implementation.
For current mission profiles, ALPHA’s red forces are handicapped with shorter range missiles and a reduced missile payload than the blue opposing forces. ALPHA also does not have airborne warning and control system (AWACS) support providing 360° long range radar coverage of the area; while blue does have AWACS. The aircraft for both teams are identical in terms of their mechanical performance.
The mission analyzed in this document features two blue fighters vs. four red [controlled by ALPHA]. The red aircraft begin over a defended coastline and the blues are 54 nautical miles due west. The blues each have 4 long range missiles (LRMs) and 4 acrobatic short range missiles (SRMs), whereas the reds have 4 medium range missiles (MRMs) onboard each platform. The reds’ radar is long range with a +/- 70° azimuth angle and a 15 degree elevation angle. The initial state of the mission is displayed below in Figure 1, with the blue AWACS off-screen, due northwest of the blue fighters.
Again, ALPHA is currently trained to utilize a force of superior numbers, but weaker capabilities. In order to complete missions safely, the blue forces must be manipulated into poor positions, as their kill range is farther than the reds’.ALPHA’s ability to defeat AI-flown enemies is only one measure of success; it must also be able to defeat highly trained and experienced fighter pilots. ALPHA was assessed by Colonel (retired) Gene “Geno” Lee. As a former United States Air Force Air Battle Manager, Geno is a United States Air Force Fighter Weapon School graduate and Adversary Tactics (Aggressor) Instructor, and has controlled or flown in thousands of air-to-air intercepts as a Ground Control Intercept officer, as a Mission Commander on AWACS, and in the cockpit of multiple fighter aircraft.
...
When Geno took manual control of the blue aircraft against the reds controlled by the baseline controller AFRL had previously been utilizing, he could easily defeat it. However, even after repeated attempts against the more mature version of ALPHA, not only could he not score a kill against it, he was shot out of the air by the reds every time after protracted engagements. He described ALPHA as “the most aggressive, responsive, dynamic and credible AI (he’s) seen-to-date.”
Esmamulje, et vennad on kokku seganud geneetilised algoritmid, hägusloogika ja otsustuspuud ning selle suht suvalisel "saial" käima vedanud. Kui see kõik on tõsi, siis on see hämmastav. Tuleb ilmselt originaalteostesse sügavamalt sukelduda.
Asjaga tegelevad pidevalt asjatundjad, lootust on. 
