在手握方向盤(pán)的同時(shí),觀察屏幕上像波浪般滾動(dòng)的α和β腦電波圖形,這無(wú)疑是汽車(chē)研發(fā)中最奇特的景象之一。但這的確是捷豹路虎正在進(jìn)行的一項(xiàng)研究——如何使用高度精細(xì)的技術(shù)手段,幫助消除駕駛員的疲勞和注意力分散對(duì)駕駛行為的負(fù)面影響。
“我們的研究項(xiàng)目名為‘頭腦感官研究’,最終目標(biāo)是研發(fā)出一種系統(tǒng),可以判斷駕駛員的精神是否集中,能否保持警覺(jué)和專(zhuān)注力,是否處于分心狀態(tài)等;這些都是通過(guò)對(duì)大腦狀態(tài)進(jìn)行監(jiān)測(cè)而得知的,”人因研究員兼項(xiàng)目主管Laura Millen博士表示。“我們不能要求駕駛員頭上頂著一個(gè)有傳感器的耳機(jī),所以還要找到一種非侵入式方式,且能有效收集數(shù)據(jù)。”
而他們經(jīng)過(guò)反復(fù)思考后,設(shè)計(jì)了這只世界上最與眾不同的方向盤(pán)。這個(gè)方向盤(pán)同時(shí)也是一個(gè)全新測(cè)試臺(tái)的一部分。Millen博士說(shuō):“我們使用方向盤(pán)背面上方的四個(gè)傳感器,對(duì)原型機(jī)的效果進(jìn)行評(píng)估。”
α和β腦波的監(jiān)測(cè)讀數(shù),顯示在測(cè)試臺(tái)前方的一個(gè)大型屏幕上。需要強(qiáng)調(diào)的是,這只是研究裝置,絕對(duì)不會(huì)出現(xiàn)在成品車(chē)的平視顯示器上。
“大腦會(huì)持續(xù)不斷地產(chǎn)生四種或更多種類(lèi)的腦電波,它們的速度和頻率各不相同。” Millen說(shuō),“盡管不同腦電波同時(shí)產(chǎn)生并相互交織的,但人在特定時(shí)段的的意識(shí)狀態(tài)是由當(dāng)?shù)闹漕l段(即最強(qiáng)頻段)的腦電波決定的。在睡眠期間,大腦的支配頻段是慢速δ波。而在做白日夢(mèng)時(shí)或?qū)⑿褧r(shí)分,支配頻段就變成了依舊很慢但比δ波略快的θ波。當(dāng)大腦處于冷靜且集中的狀態(tài)時(shí)——比如當(dāng)一個(gè)人正在閉眼放松時(shí),支配腦電波就換成了比θ波更快的α波。”
為了測(cè)試駕駛員的注意力集中程度,研究人員讓他進(jìn)行虛擬駕駛,在屏幕上顯示的是一條虛擬的蜿蜒道路,此時(shí)研究人員使用一個(gè)分析系統(tǒng),為駕駛員的α和β腦電波“打分”,分?jǐn)?shù)范圍為0-100分。
還有許多問(wèn)題需要解決,其中包括如何警告駕駛員(可能是通過(guò)方向盤(pán)或踏板震動(dòng)的方式)。在未來(lái),該項(xiàng)目還將吸收神經(jīng)學(xué)家參與。
捷豹路虎研發(fā)主管Wolfgang Epple表示:“即使駕駛員的眼睛盯著路面,但如果注意力不集中,或處于白日夢(mèng)的狀態(tài),也意味著他們并沒(méi)有專(zhuān)注在駕駛?cè)蝿?wù)上。”
除了駕駛員的精神狀態(tài)外,研究人員還監(jiān)控他們的身體健康狀態(tài)。捷豹路虎正在研發(fā)一種“健康座椅”,可通過(guò)監(jiān)測(cè)心率和呼吸頻率來(lái)檢測(cè)駕駛員的緊張程度,以及是否存在嚴(yán)重的健康問(wèn)題。在研究過(guò)程中,駕駛員的心率和呼吸頻率都會(huì)顯示在屏幕上。同樣,對(duì)駕駛員進(jìn)行身體狀態(tài)提醒的方式仍有待進(jìn)一步研究,而且不能與喚醒注意力的提醒方式相沖突。
在飛機(jī)駕駛艙和飛行甲板上使用的語(yǔ)音提示可能是一種選擇,但捷豹路虎的研究人員并未對(duì)此進(jìn)行評(píng)論。
捷豹路虎的研發(fā)范圍非常廣泛,其中自然包括“自動(dòng)駕駛汽車(chē)”。Epple表示,比起“無(wú)人駕駛汽車(chē)”這個(gè)感性的表述,捷豹路虎更喜歡“自動(dòng)駕駛汽車(chē)”這個(gè)名稱(chēng)。“我們希望能讓駕駛員在自主駕駛和自動(dòng)駕駛之間自由選擇。這意味著我們最終研發(fā)出來(lái)的車(chē)輛可以按照駕駛員的要求進(jìn)行自動(dòng)駕駛,但它同時(shí)擁有一套智能系統(tǒng),可以讓駕駛員更多參與駕駛過(guò)程,使車(chē)輛行駛更加安全。”
捷豹路虎的先進(jìn)駕駛員輔助系統(tǒng)(ADAS)使自動(dòng)駕駛成為可能,并使“真實(shí)”的駕駛體驗(yàn)更加安全愉快。
ADAS的輔助功能包括主動(dòng)巡航控制、車(chē)道保持、自動(dòng)緊急剎車(chē)等系統(tǒng),它們可以在無(wú)需駕駛員參與的情況下完成多點(diǎn)180°大轉(zhuǎn)彎。此外還有一個(gè)駕駛員離車(chē)自動(dòng)駕駛系統(tǒng)可以勝任路虎Land Rover車(chē)型99%的非公路駕駛?cè)蝿?wù)。這個(gè)系統(tǒng)可以讓駕駛員在車(chē)外也能控制陷入困境的車(chē)輛,使其緩慢前移,而這一切只需在智能手機(jī)上進(jìn)行簡(jiǎn)單的操作即可完成。記者在捷豹路虎的Gaydon研發(fā)中心以乘客的身份體驗(yàn)了這項(xiàng)功能,他坐在車(chē)內(nèi),但身邊沒(méi)有駕駛員操縱車(chē)輛。車(chē)輛在一只前輪騰空的狀態(tài)下越過(guò)了一個(gè)攻角特別高的障礙物,這個(gè)過(guò)程非常有意思。
此外,該系統(tǒng)可以用于捷豹路虎的任何一款車(chē)型,可以幫助車(chē)主從擁擠的停車(chē)場(chǎng)中將車(chē)輛遠(yuǎn)程提取出來(lái)。
Epple認(rèn)為,該技術(shù)與遠(yuǎn)程停車(chē)控制技術(shù)是相互關(guān)聯(lián),共同演進(jìn)的,它們將極大地提升公眾對(duì)自動(dòng)駕駛汽車(chē)的信任度,更公眾容易接受這一新概念。Epple將這種技術(shù)稱(chēng)為“車(chē)輛自動(dòng)前進(jìn)技術(shù)。”
他認(rèn)為“一定程度的”自動(dòng)駕駛技術(shù)將在2020年后不久進(jìn)入市場(chǎng),最初將用于中間有隔離帶的雙幅路,隨后逐漸用于更窄的道路,而捷豹路虎的全自動(dòng)駕駛汽車(chē)將于2025年面世。
雷達(dá)、激光雷達(dá)(LIDAR)和立體攝像機(jī)將為自動(dòng)技術(shù)提供支持。Epple表示,捷豹路虎要求“感知、運(yùn)動(dòng)規(guī)劃和導(dǎo)航/定位”這三種功能具備不同程度的認(rèn)知能力。這意味著捷豹路虎汽車(chē)的自動(dòng)駕駛功能不僅能在沒(méi)有路面警示標(biāo)志的道路上使用,而且還適用于完全沒(méi)有道路基礎(chǔ)設(shè)施的情況,包括在沙漠中。
這個(gè)解決方案已經(jīng)在捷豹路虎的自動(dòng)駕駛汽車(chē)Solo Car上有所運(yùn)用,并將降低成本的激光雷達(dá)傳感器(捷豹路虎的研發(fā)項(xiàng)目之一)和一系列全新的傳感器進(jìn)行了有效結(jié)合。然而在研發(fā)新傳感器的同時(shí),捷豹路虎也在努力加強(qiáng)現(xiàn)有傳感器的功能,公司內(nèi)部將這一努力稱(chēng)為“物盡其用”。
“運(yùn)動(dòng)規(guī)劃體現(xiàn)出車(chē)輛對(duì)以下信息的了解——車(chē)輛應(yīng)隨時(shí)了解自身的位置、即將前往的目的地,以及如何到達(dá)目的地。”Epple解釋道,“導(dǎo)航/定位則反映了車(chē)輛在行駛中的導(dǎo)航能力,以及目前正行駛在整個(gè)路段的哪個(gè)部分。”
在駕駛過(guò)程接近尾聲的時(shí)候,駕駛主動(dòng)權(quán)將交還給司機(jī)。此時(shí)車(chē)輛會(huì)檢查駕駛員是否清醒,注意力是否集中,這就是“頭腦感官技術(shù)”和“健康座椅技術(shù)”發(fā)揮作用的時(shí)候了。
正在研發(fā)中的其他駕駛員提醒系統(tǒng)還包括觸覺(jué)式油門(mén)踏板,當(dāng)騎車(chē)人或行人接近汽車(chē)的時(shí)候,踏板會(huì)通過(guò)振動(dòng)或加大扭矩的方式來(lái)提醒駕駛員?!镀?chē)工程雜志》的編輯發(fā)現(xiàn),踏板震動(dòng)提醒很有效,但增大扭矩會(huì)令人不安。因?yàn)樵谝恍┨囟ㄇ闆r下,加速就如同剎車(chē)一樣重要,而在緊要關(guān)頭卻需要駕駛員用更大的力氣才能踩動(dòng)油門(mén),這并不合理。
此外,捷豹路虎還在研發(fā)一種預(yù)測(cè)式觸覺(jué)設(shè)備,可以跟蹤車(chē)載信息娛樂(lè)系統(tǒng)屏幕前方的手部動(dòng)作,來(lái)預(yù)測(cè)駕駛員將按下哪個(gè)按鈕。該技術(shù)可以將駕駛員“視線(xiàn)離開(kāi)駕駛前方”的時(shí)間減少五分之一左右。在靜止?fàn)顟B(tài)時(shí)的演示結(jié)果很有說(shuō)服力,但當(dāng)車(chē)輛行駛在顛簸路面上行駛——尤其是非公路駕駛工況下,駕駛員的肢體動(dòng)作無(wú)法預(yù)測(cè)的時(shí)候,該系統(tǒng)的效果還需進(jìn)一步驗(yàn)證。
捷豹路虎的其他研發(fā)項(xiàng)目還包括汽車(chē)自我學(xué)習(xí)能力(公司聲稱(chēng)這是一項(xiàng)“全球首創(chuàng)”的研發(fā)項(xiàng)目),該能力可以“了解”駕駛員上下班時(shí)最喜歡聽(tīng)的廣播頻道,以及在什么時(shí)候座椅按摩,需要按摩力度需要多強(qiáng)。
盡管捷豹路虎的科技研發(fā)越來(lái)越全面,但并未打算讓車(chē)輛取代駕駛員。Epple笑稱(chēng):“車(chē)上坐的是駕駛員,不會(huì)變成貨物!”
作者:Stuart Birch
來(lái)源:SAE《汽車(chē)工程雜志》
JLR reveals some R&D secrets
Watching the alpha and beta waves of one’s brain undulating across a screen while gripping a steering wheel, is one of the more bizarre experiences of automotive R&D. But that is what JLR (Jaguar Land Rover) is putting its mind to as it delves deep into how to eliminate the effects of driver fatigue or inattention using highly subtle techniques.
“Our project is called Mind Sense and has the ultimate goal of developing a system to detect whether a driver is concentrating, alert and focused, or distracted; we are doing so by monitoring what is going on in the brain,” said Dr. Laura Millen, Human Factors Researcher and leader of the project. “We can’t expect drivers to put on a headset carrying sensors so part of our research project is to look for non-intrusive ways to collect the same data.”
The result is what is probably the world’s most extraordinary steering wheel, which forms part of a novel test rig: “We are evaluating a prototype with four sensors on the back of the upper part of the wheel’s rim.”
Readings are presented on a large screen (this is for research only and definitely not something that would appear on a production car’s head-up display!) ahead of the rig, indicating alpha and beta brainwaves.
"The brain continuously produces four or more distinct speeds or frequencies of brain waves," explained Millen. "Although these different brain waves are produced simultaneously and in combination, a person’s state of consciousness depends on the dominant (strongest) frequency band at each time. During sleep, the brain produces dominant slow delta waves. During daydreaming or in the twilight of sleep, the brain produces dominant theta waves that are slow but a bit faster than delta. When the brain is calm and mentally unfocused—for example, when a person relaxes with their eyes closed—the still faster alpha waves are dominant."
To test levels of concentration, the alpha and beta waves are “scored” by an analysis system on a zero to 100% scale as the rig driver watches a screen showing a virtual drive on a winding road.
There are many questions to be resolved, including how to warn a driver (possibly steering wheel or pedal vibration), and later the project will embrace inputs from neuroscientists.
JLR R&D Director, Dr. Wolfgang Epple, said: “Even if the driver’s eyes are on the road, a lack of concentration or a daydream will mean he or she isn’t paying attention to the driving task.”
Complementing this research into a driver’s mental state is the monitoring of his or her physical health. JLR is developing a “wellness seat” that can detect stress, or some serious health issues, via heart- and respiration-rate monitoring. For research, heartbeats and breathing rates appear on displays. Again, alerting methods that don’t conflict with other attention-getters need to be refined.
Voice alerts, such as those used in cockpits/flight decks for aircrew, would seem to be a possible solution, but JLR’s researchers did not comment on this option.
JLR’s R&D operation spreads very wide and not surprisingly covers the “autonomous car,” a phrase Epple said the company prefers to the emotive “driverless car”: “Our vision is to offer a choice of an engaged or autonomous drive. Ultimately, this means that a car could drive itself if the driver so chose, and have intelligent systems that can be adjusted for a more engaging, involved, and safer drive.”
The company’s Advanced Driver Assistance Systems (ADAS) will enable autonomous driving and make “real” driving safer and more enjoyable, he explained.
Supporting this, together with active cruise control, lane keeping, autonomous emergency braking and other systems will be the facility to make a multi-point 180° turnaround without driver input, and an autonomous, driver-out-of-car facility that could cope with 99% of a Land Rover product’s off-road capability. This would enable a driver tackling very difficult situations to inch forward a vehicle remotely by use of a mobile phone with simple controls. Experiencing this at JLR’s Gaydon R&D center as a Range Rover passenger with no driver aboard, progressing over an obstacle to a very high angle of attack with a front wheel in the air, proved interesting.
The system could also be used by any JLR product to remotely extract a car from a hemmed-in parking slot.
Epple believes such technology will prove a major aspect of public acceptance of autonomous vehicles because it will engender trust, much as park distance controls have done. They are all part of an integrated whole and an evolution of technologies that he defines as “autonomous progression.”
He sees “a degree” of autonomy entering the market soon after 2020 for use on dual-carriageways, followed progressively by lesser roads, and with a JLR fully autonomous vehicle a reality around 2025.
This development will be supported by radar, LIDAR, and stereo cameras. Epple stated that JLR required levels of cognitive ability in three components: perception, motion planning, and navigation/localization. This would give a JLR vehicle autonomous capability not only on roads without surface warning markers but also without roadside infrastructure including deserts.
Such a solution has been dubbed by JLR the Solo Car, able to take care of itself and incorporating reduced-cost LIDAR sensors (JLR has a program to achieve this) working with a range of new and different sensors. However, where possible JLR is extrapolating the capability of existing sensors, known within the company as “sweating the assets.”
“Motion planning describes the vehicle’s understanding of where it is at any given point, where it has to get to, and what it needs to achieve that,” explained Epple. “Navigation/localization describes the vehicle’s ability to navigate to the destination but also understand on which part of the journey it is currently traveling.”
In handing over to a driver at completion of what may be a lengthy autonomous journey phase, a vehicle could check that the driver is awake and paying attention, which is where Mind Sense and the Wellness Seat enter the equation.
Other R&D systems to support or alert the driver include haptic accelerator pedals that will vibrate or need added torque to operate, to get attention for such things as cyclists or pedestrians entering the car’s vicinity. This Automotive Engineeringeditor found the pedal vibration warning technique effective but added torque disconcerting; there are situations when accelerating can be as necessary as braking, and a throttle pedal requiring an extra shove in a looming potential emergency was unconvincing.
JLR has also developed a predictive touch prototype that tracks hand movements in front of an infotainment screen to predict which button a driver will press. This is to reduce “eyes out of cockpit” times by about a fifth. A static demonstration was convincing but a demonstration was required of the system in a car on a poor surface with the vehicle experiencing unpredictable body movements, particularly off road.
Other JLR programs include the self-learning car (claimed as a “world first”) that can “know” what radio stations are a driver’s favorite when driving to and from a workplace and what level of driving seat massage is required and when.
However, though JLR works towards ever more technologically comprehensive products, it has no plans to replace the driver, quipped Epple: “The occupants of our cars will not become cargo!”
Author: Stuart Birch
Source: SAE Automotive Engineering Magazine