集微访谈 |Ahmed Banafa:5G、AI、元宇宙等爆发式发展,智械元年来了吗?

作者: 陈炳倩 2023-03-14
AI解读文章
来源:爱集微 #集微访谈# #物联网# #元宇宙#
1.7w

集微网消息,在往期的集微访谈栏目中,爱集微有幸采访了圣何塞州立大学讲师兼学术顾问Ahmed Banafa。他在专注于物联网、区块链、网络安全以及人工智能,其研究被世贸组织和亚太经合等国际组织引用。集微访谈就物联网、元宇宙人工智能相关方面提出了一系列问题,并收到了十分有启发的答复。

问:目前的物联网设备似乎使用不同的网络通信标准,例如蓝牙、WiFi、ZigBee和4G LTE。为什么会这样?

答:具体来说,这是开始物联网通讯部分之前的事情 。物联网由四个部分组成,第一部分是传感器,第二部分是网络,第三部分是云/AI,最后一部分是应用。我是这样向我的学生解释的。所以在谈论通讯,即此处的网络,如 5G。5G 是一种可能的通讯,和 NB-IoT 是同样的东西。美国已开始使用基于 4G 的 NB-IoT,但现在将它归为 5G 的一部分,实际上已经包含在5G里面了。这只是一种选择,比如你有一台设备,它并不是只能用一种网络。它可以使用 Wi-Fi,也可以使用蓝牙,使用另一种网络,LTE,随便什么。这取决于设备或传感器的可靠性。我正使用耳机听音乐,耳机通过蓝牙连接电脑。而我的麦克风则是使用有线方式连接 ,此外我还有 Wi-Fi(把电脑连至网络)。只有一种通讯方式不会有任何帮助。你需要有多种选择,根据哪个速度更快、哪个适合处理这个负载,从一种切换到另一种。在做某些事,如下载应用程序,或下载文件时,也可以看到这一点,他们会等到连接 Wi-Fi 才进行,因为这比使用手机网络更快。

问:未来物联网设备使用的网络通信标准是统一还是多种方式并存?

答:这就像诸如 WiFi6、6G 或者是人们可以使用的某种卫星互联网之类的东西。我的意思是,为一台设备制定多个标准并没有什么错,这实际上会给设备带来更多的可用性,因为世界上不是每个国家都有相同的网络功能。来到一个国家会发现他们只有 4G 或 3G。若设备无法正常工作,那么触及到的市场就更少。设备要兼容多种标准,这就是所有设备都有这个备用技术选项的原因。如果我购买的东西,该设备可用之前的通信版本工作,这取决于销售此产品的地方。

问:现在,数字主权。它的主权是国家安全日益增加,越来越成为各国最关心的问题之一。全球数据共享变得越来越困难。那么这是不是谷歌物联网标准化进展缓慢的重要原因呢?答:这个问题不错。物联网的标准化是物联网普及的最大障碍之一,稍后我会回答这个问题,首先举个例子,如果我买了一台设备,我不必担心它是否支持连接到Alexa或Siri。它应该在任何地方都能连接——这就是标准化。这可以通过所谓的规范来解决,来约束每个制造商,它们都会有一个合规部门来回答例如GDPR之类的问题,你也听说过GDPR(通用数据保护条例)。他们必须对此作出回应,以确保公民的隐私、安全真正得到保护。一旦我们解决了各国的担忧,这些设备就很容易被购买和销售到这些国家。如果没有规范,批准这些设备上市就没那么容易了。这就是为什么在这种情况下,规范很重要,而标准化必须由非营利组织来完成,就像IEEE为我们带来了WiFi和其他标准。他们可以为我们带来物联网的某种标准化。

问:未来可穿戴设备与AI的结合会有哪些突破?

答:回到物联网模型,我们谈论四个组成部分传感器,通信网络/AI应用程序。现在讨论第三个要素,也就是AI,当你获得数据的时候,你需要想办法进行处理。而当你想处理数据的时候就需要一种方法来实现,而且以一种非常智能的方式来做不管处理流程是在设备上还是从边缘设备发送到云端。现在我告诉我的学生,当我告诉他们,听到“智能”这个词,智能电视,智能汽车,智能手机,须明白“智能”这个词意味着三件事。其一是操作系统,通常是安卓系统。其二 IP 地址,用于连接互联网。其三是算法,也就是AI。如要说这个麦克风是智能的,必须有三个部分。物联网设备同样如此。若有智能相机,相机会有一些算法,能够进行小规模的处理。比如相机识别到一张特定的人脸,而这个人脸和之前设定需要警告其他人的人脸一致时,或者在采集场景检测到了有东西在运动时,相机需要决定是否应该启动。所以这一个简单的开关逻辑,而是需要更智能的处理方式。这就是人工智能存在于边缘设备之中,同时人工智能也存在于云上,可以先采集上云再处理对物联网网络中发生的事件能识别得更好

可穿戴设备是一股即将到来的浪潮。例如,Apple Watch其他的智能手表通常它用于健身检查脉搏情况,但它们能做到的不止这些。现在可以看到很多时候手表能用来执行一些所谓的快速任务不需要拿出手机,只看手表可以快速地处理一些事情,因为手表和其他设备连接的,如果有信息,有事情,或者预约用手表处理这对我来说容易点。无线可穿戴设备将不仅仅是只是智能手表。它可能是我衣服上的东西,例如,可能是我的眼镜,比如说它可以帮助检查我的眼睛,可以检查我有什么健康问题很多功能目前都是需要人工的,你得等结果出来。此信息其实可以直接上传并发送医生。例如,有一个技术人员去检查其中一个管道或其中工厂,他或她可以直接点击可穿戴设备,它会弹出一个菜单,通过它可以看到被检查设备的状态。可穿戴设备将为虚拟现实和增强现实提供更多的驱动力和推动力,这是可穿戴设备的应用之一可以成为世界的一部分。所以我把它称为群体、技术的第三维度,因为现在谈论的技术正在朝着一个没有想过的方向发展。技术已成为你的一部分,你可以带它到任何地方,甚至不在你的口袋里。我写了一篇与此相关文章实际上,在某些情况下,你的皮肤下可能嵌入有芯片。这个芯片上会有你的所有信息可以用它做任何事情可以用它来付钱,进入大楼,获取信息,了解你的健康信息一切事情不需要用你的手机和手表个芯片将是你的身份证明。所以可穿戴设备是下一次革命,而且我们很快就会看到。

问:如何解释“软件定义汽车”这个术语?

答:软件定义汽车是几年前开始使用的通用术语,基本上是“基于软件的硬件”的概念,在软件定义车辆的情况下,有车辆,有软件层的堆栈,然后有司机车辆与司机之间是软件层。现在关于此重要的是,看所有基于软件和程序的汽车时,基本上是一个带四个轮子的电脑,就是这么回事。主要集中在硬件和软件上。在这种情况下,我的是电动汽车。从技术上讲,它的其余部分并不像传统汽车那样复杂。它不需要那么多机械部件。这很好,为什么?因为有一点,如果想修复召回,可以通过 OTA 来做到,这可以在特斯拉一直看到这一点。如果有召回,不必去找经销商什么都不用做他们可以向您发送信息来解决问题。软件定义车辆指随着时间的推移,软件的比例开始扩大,先是成为汽车的一个简单组成部分,然后一点点变大,达到汽车设计总量的 50% 。这种情况下一辆车会同时有机械电气和软件三类部件。在我谈到软件定义车辆时,我想到的术语是方便、舒适以及安全,这里真正困扰我的安全。因为汽车是基于软件,若有人入侵会发生什么。但另一方面,由于它是软件,而软件可以向制造商传输数据。从制造商的角度他们得到了大量的信息这将有助于完善他们的产品。不通过从大量汽车中选择或调查或通过一些旧的传统方法来收集信息,他们有大量的数据传输到他们的服务器可以使用大数据分析,看看对于司机来说汽车的哪些功能有问题,缺少什么功能,可以添加什么功能。这将打开升级的大门,这基本上是更多的钱,提供更多的功能,确保要在下一个版本中更改哪些功能所以当通过软件定义汽车的时候,制造商和司机联系是从买这辆车的那天开始而不是结束。

谈论智能时,你谈论的是让驾驶不那么紧张和危险的功能因为在这种情况下,汽车可以相互交流,汽车会知道哪里发生了车祸,可以交换某辆车和另一辆车的信息目前缺少的那块拼图是智能道路,这是我们接下来要做的。比如,智能道路有传感器,可以为电车充电。同时,还可以给汽车发送信息,告知路况。所以要拥有这种完美的功能,必须有智能汽车智能手机智能道路,还有智能的“司机”。想要智能的“司机”那么得先有自动驾驶汽车,也就是“司机”。在旧金山,我们有一家名为Cruise的公司,经营无人驾驶出租。有时会出现一些小故障,例如,上个月有辆车直接停在车流中间,在那之前2个月警察拦下了车,因为这辆车停在不该停的地方。警察靠近那辆车时,开始加速并离开现场。所以像这样的事情必须解决的。目前我们还在一个非常早期的阶段,正如我之前说,智能汽车是物联网的一部分。智能手机是每个人一生中拥有的第一个物联网设备。所以为了完成整个构想,必须要凑齐这四个组成部分,任何地方都要有传感器,到处都有信号可供连接,能处理数据的云计算最后还需要为这些场景准备的应用程序。

问:有篇文章形容现在的智能汽车系统可以让你感觉随身携带了四个轮子的智能手机,那么智能手机和智能汽车将会是竞争关系么?

答:这个问题不错。它们会融合。我们看到融合正在进行。很多时候你会发现,你买了一辆车,他们会问你想用iOS还是安卓系统?一旦你选择了操作系统,它就会把汽车的所有界面和功能都切换到该操作系统。我的汽车有这种选项:可以选择使用安卓或iOS手机连接,这基本上是一样的。我的手机就在我身边。然后我可以在车机屏幕中看到我的手机屏幕 。因为车机的屏幕更大,所以我可以看地图,我可以打电话,我可以看到短信——所有这些功能可以在车机上,所以它们会融合,相互支持。这就是我的看法。你知道有些汽车你可以用手机打开车门。这是两技术如何融合的一个例子。然后制造公司汽车行业明白每个人的手中都有手机。他们会确保这是他们可以利用切入点。所以他们可以在手机上提供关于汽车功能。

问:边缘计算有什么优劣?

答:所以要理解什么是边缘计算,边缘计算是将算力尽可能靠近事件,这基本上类似于你可以在各个地方找到小型数据中心。他们研究市场,看哪里有很多流量大量的用户,不是把所有东西都放在不同州甚至不同的国家的大型数据中心里。你需要将计算靠近用户。这样可以减少延迟。这样延迟就会很低。带宽本身将节约下来。为什么因为只是为了让我关掉一盏灯,我不需要通过互联网,把指令传输到千里之外的另一个,然后再把指令发送给我房间的灯。而是会有一个区域,在区域内有一个小型数据中心,距离设备用 很近。这就是它的美妙之处。这就是为什么边缘计算是一项受到许多公司关注的技术,包括英特尔、戴尔、微软谷歌公司。他们想要这项技术,因为这样可以节省带宽。不是所有人都使用同样大的带宽相同的大型数据中心,同时,这将解放数据中心的时间来做计算。到一天结束时,边缘计算设备中心处理的所有数据都可以上传到大数据中心进行处理。

这是一个正确的方向,你需要这项技术。我也给你举个例子。特斯拉是推动边缘计算的最大公司之一,因为他们的汽车必须在瞬间做出决定。如果你要这样做,那么你不能等到数据到达数据中心再到你这里。但如果附近有一个边缘中心或边缘计算数据中心,通信就会更快。所以我们可以做出决定。Netflix一直在使用边缘计算,根据有多少人在观看某一电影。因此,全国将不再有20个大型数据中心,而是数百个小型数据中心,它们可以非常快速地处理这些指令。这就是两者的区别

问:相对于在民用市场不温不火,数字孪生在工业领域倒有比较明确的用例,是什么导致了这两者的差异?

答:10年,到2030年。到那时,你会发现虚拟现实、增强现实、扩展现实和元宇宙真的会成为某种商 品,人们可以用它来工作、上学、接受采访和办公。但这需要时间。因为你需要很多东西才能实现这一点。例如,我们现在的采访是二维的。如果我们有了元宇宙,我会有一个虚拟化身代表我,你也会有一个虚拟化身代表你,我们会坐在彼此面前。我们可以用虚拟化身来交谈,我们可以握手,我可以给你看一些东西,你可以给我看一些东西,我可以设计我的办公室。我不需要虚拟背景,它会建立一个背景。根据马克·扎克伯格和元宇宙从业者的说法,大约需要10年。所以到2030年,他们会看到元宇宙的某种应用。我们已经有一些元宇宙,你可以去参观。有一个facebook horizon。那是他们自己的元宇宙,人们可以四处走动,互相了解,坐下来谈论买房,举办音乐会,但元宇宙仍处于早期阶段。所以在10年内,它会向我们展示所有这些应用,但还为时过早。

我的意思是,美国有些大学有一两门元宇宙课。基本上是生物或化学课,他们展示如何做事。但这始终是开始。没有人能保证它会成功,但你看到了才会知道。到目前为止,META 花费了大约130亿美元,仅仅是为了拥有基础设施和进行所有相关研究。

没人说明天或2025年就能实现。在这10年的尺度里,我们会看到它是如何普及的。一旦你看到人们谈论居家办公的应用,元宇宙就是一种解决方案,你可以居家办公。基本上,你可以在办公室里,人们可以看到你,你可以有你的虚拟化身。所以”可以坐在办公室里工作,而不是坐在家里躲在摄像机后面。也许还有未开发的应用场景,但我认为除了游戏之外,它不能为消费者提供其它应用。正如你所说,对于这个行业来说,数字孪生是非常有帮助的,为他们中的许多人、制造业和公司节省了资金和时间。

直到我们从制造业或者从商业和工业领域取得这一成果,这将把我们带到工业物联网。从工业物联网发展到物联网或者消费物联网。

都投在了工业物联网。工业物联网钱比什么都重要,因为在市场和物联网中,你的客户数量有限,但他们有很多钱。当你谈论消费者物联网时,会有很多令人头疼的事情,因为你要面对数十亿的消费者。你有各种不同的法律法规隐私问题,所有这些你必须处理。

以下是采访原文(英文):

Q:It seems that current IoT devices use different Network communication standards, such as Bluetooth, WiFi, zigbee, and 4G LTE. Why does this happen?

A:That is the thing before we start. getting into specifically the communication part of the internet of things. And the thing is four parts, four components. The first one is the sensor, the second one is the network, the third one is the cloud slash AI and the last one is their application. This is, this is how I explain it to my students. So when you talk about the communication, the network here, for example, like the five G. Five G is one of the possible communication, a narrow b and IUT is the same thing. Now, here in the United States, they started the narrow band NB-IoT they started with, based on four G but now they have it as part to five G. So it’s actually included there. It’s just an option, because when you have your device, for example, you could be running from one communication network to the other. You could be going to the Wi-Fi. And you can go to the Bluetooth and you can go to another network, LIT [], whatever. it depends on the accountability of the device or the sensors. I’m listening to using my headsets and this connected to the computer with the Bluetooth. I have my microphone connected to wired and I have a wireless Wi-Fi. So the having just one kind of communication is not goanna help. you need to have multiple options when you switch from one to another based on which one is faster, which one will process that for you. And you can see this also when you are doing something like what’s app or when you are downloading files, they wait until they get to the wifi because that can be faster instead of using your cell phone network.

Q:In the future, will the Network communication standards used by IoT devices be unified or will multiple methods coexist?

A:It could be something like wifi, six, or six g for example, or some kind of a satellite internet that people can use for. I mean there is nothing wrong of having multiple standards for one devices actually give the device more availability because not every country in the world will have the same capabilities of the network. We go to a different country. You’re gonna find that they only have four g or they have three g if your device is not working, your market will be less. You need your device to be available for multiple standards. And this is why this back technology option we have it in all the devices. If I buy something that device can work with the previous versions, because you know it depends on where you are selling this product.

Q:Now, digital sovereignty. it's a sovereignty is that national security are increasing, become increasingly becoming one of the top concerns of all countries. And global data sharing is becoming the increasing difficult. So is this an important reason for slow the progress of google, iot standardization?

A:A good question. IoT standardization is one of the most difficult obstacles for IoT to spread everywhere, and then i'm gonna answer that question. For example, if I buy a device, I don't have to worry that this is gonna be connected to alexa or siri. It should be connected to everywhere. So that's the standardization. This is can be solved with something called compliance. Every single manufacturer will have a compliance department that will answer the questions for something like gdpr and you heard about the gdpr the general data protection regulations.

They have to answer to that to make sure that privacy, security, and safety is really preserved for the citizens. Once we answer the concerns of the countries, it's very easy for the devices to be purchase and to be sold to other countries. Without it is gonna be very difficult for them to approve it. And this is why compliance is important in this case, and the standardization has to be done by nonprofit organization, like itreablry brought us so many of those standards like the wifi and the others. They can bring us some kind of standardization for the IoT.

Q:What breakthrough will be made in the combination of wearable devices and AI in the future?

A:You know, back to the model of IOT when we talk about the four components, the sensors, the communication, the cloud slash ai and then the application. Now we are talking about the third element, which is the AI you get this information and you process it. When you process it, you need one way for you to do that, and you do it in a very intelligent way whether it is on the device itself. Or he can send it to an edge,you know, device or he can send it to the cloud. Now I tell my student, when I told them, whenever you hear the term smart, smart TV, smart car, smart phone, you have to understand that the term smart means three things. Number one is an operating system, which is usually in android. Number two is IP address, which is connected to the internet. Number three is algorithm, which is AI. So in order for me to say that this microphone is smart, I have to have the three components. The same thing gonna go with the IOT devices. When we have a smart camera, that camera, they will have some algorithm that will process. I may be on a small scale, should I turn the camera on if this face is one of those faces that I have to alert somebody that because I recognize the face, or maybe there is a movement in the area, it’s not gonna be like a simple on and off. You need something more intelligent. And this is where you have the artificial intelligence exist on this the edge devices and also exist in the cloud to process the data later and try to have a better understanding of what’s going on in the network.

OK. So, wearable devices is a wave that’s coming very soon. For example, if we talk about the I watch or the smart watches, usually it was used for, you know, fitness just to check your pulse. And that’s gonna be more. Now, you will see that many times they call them the short task, which is instead of me to pull out my phone, I can just look at my watch and then I can do the things very quickly, because that one is connected and it’s much easier for me if there is a message, if there is something, if there is an appointment. So they use for the wireless for the wearable device is goanna be more than just having it as a watch. It could be something with my clothes, for example, that could be somebody with, you know,  you know, my eyeglasses, for the glasses, for example, that could be a something that will help me can check my eyes. They can check what’s wrong to me. So many of the task that has been done manually now and you have to wait until you get the result. This information can be uploaded and sent to your doctor, you know, can be sent to, for example, if you have a technician and go and check one of those pipelines or one of the factories, and he or she can just hit this one. And there is a menu. And they can just go through that and see what’s going on. It’s, it’s the wearable communication, wearable devices will give more, give more drive and push for the VR and the AR, the virtual reality and augmented reality, because that’s one of the applications of the wearable something you wear, and you can be part of the world. So this is I call it the third dimension for, you know for a community, for technology, because we are talking about, you know technology now going in the direction where you are not going to a place for the technology. The technology is part of you, you can carry it with you everywhere, is not even in your pocket. It’s actually, in some instances, and I wrote an article about this. You could have a chip inside under your skin. And this chip will have all the information about you. You can use it for everything. You can use it for paying, for accessing buildings, for information, about your health, for everything, instead of having your phone and having your watch, that’s goanna be your identification. So wearable is the next revolution, and we’re goanna see a lot coming soon.

Q:how to explain the term "Software Defined Vehicles"?

A:Software define X is the general term that started years ago, which is basically the, you’re talking about conceptual of a hardware using software. in the case of the software defined the vehicle, you have the vehicle, and you have the stack of the layer of the software and then you have the driver. So, between the car and the driver is the layer of software. The important thing about this is now, when we look at all those cars that is built based on software and programming, it’s basically,A computer has four wheels. That’s what it is. It’s mostly focused on the hardware and the software. There is now and I’m talking about electrical cars. In this case, the rest of it is, technically speaking, is not as complicated as what we have in the traditional cars. You’re not talking about a mechanical component here. It’s great. Why? Because here’s one thing about, if you would like to fix a recall, you can have something over the air update. You know, you can do something over the air update, and we’ve seen this one in tesla all the time. There is a recall, you don’t have to go to the dealership. You don’t have to do anything. They can just fix it for you by sending you this kind of information. It’s the concept of software defined vehicles, it is the percentage of software over time start expanding to use to be a simple part of the cars and start expanding spending and hitting,the 50 % of the design of the car. In this case, you have mechanical or you have electrical and software at the same time. When I talk about the software defined vehicle, it just that the terms that comes to my mind is convenience, comfort, you know, We talking about safety. In this case, the things that really bothering me is security. What happen if somebody hacked into. Because now it’s a software once you have it as a software and you communicate with the manufacturers. From the side of the manufacturers. They get these tons of information. And that’s goanna improves their products. Instead of it gathering this one through a selective number of cars or survey or doing some of the old traditional ways. They have tons of data transmitted to their servers. And they can just use big data analytics and see which feature the drivers have a problem with, what is missing, what can we add? And this will open the door for upgrades, which is basically more money, providing more feature people ensure that this is what you need to be changing in the next version.So you go to software. Your relationship with the driver starts the day. You buy the car. It’s not the end.

When you talk about smart, you mean you’re talking about many of the features that goanna makes driving less hectic and less, you know, danger because cars can communicate with each other in this case. So, they will know that there is an accident. They can exchange information about something going on what car compared to the other car. What is missing in the picture is smart roads that will be the next way for us. Those smart roads will have sensors to charge their cars, for example, of their electrical. At the same time, it’s goanna start emitting information to the cars, telling them what is the issues of the road. So to have this perfect feature, you have to have smart cars, you have to have smartphones, you have to have smart road, and also you have to have smart drivers. And to get rid of the smart drivers, you can have the autonomous cars which is the drivers. In San Francisco, we have a company, that the name of the company is Cruise, that one is running a driver’s taxi. And it has some glitches from time to time. For example, last month is stopped in the middle of the traffic and 2 months before that police stop the car because the car was parking in a place you shouldn’t be there. When the policeman approached the car, the cars start accelerating and left the scene. So things like this, which is has to be how to be fixed there. This is a very early stage, because as I said before, smart cars is part of the internet of things. Smartphones is the first IOT device anybody have in their life. You know, so in order for you to complete the whole picture, you have to have the four components. We have to have sensors everywhere. We have to have capability of communication everywhere. We have to have processing in the cloud, and we have to have apps for so many of them.

Q:And I have read some articles and that article said that because the smartphone has more universal than the smart cars because the cars is you have already said the class is the computer with four wheels does the smartphone we can say that it's the computer on your hand. So you can take it everywhere. Somewhere the smartphone as my smart car didn't, they can go as in your house.

So some . of view is that the future contractors, the major company of smart cars, is the phone in your hand, because when you're in the car, you can use the smartphone for your entertainment, for the navigations, for anything, the smart features. But your car can only have that features when you physically in your car. When you're out of your car, you didn't have.

So some consumers prefer to buy some the to buy some the video theories, the video descriptions, the term net flicks, the audio apple music does that. Then should like that. The description on their phone is not on or by applications on their phones, the smartphones rather than the smart cars. So what do you think of this phone review? Because in some , the smart cars and the smartphone are the compact can competitors.

A:That's a good point. They're gonna merge. And we see that it's gonna a matter of merging. Many times you will find, you will buy a car, and they will ask you, would you like to go with an ios or they're gonna go the android? Once you pick the operating system, it's gonna switch all the interface and all the features of the car to that operating system.

In my car, I have two options. They have to go ios or android, and it's basically the same thing. So my phone will be next to me. And then I can see the screen of my phone on the car, on the car screen and a bigger screen. So I can see the map, I can make phone calls, I can see the text messages. All of these things will be available, so they will merge as one helping the other. That's how I look at it. You know that some of the cars you can open them with your phone, you can open the door with your phone. That's an example of how two technologies are emerging. Then the manufacturing company, the car industry understand that phone is in every hands. They will make sure that this is an access point they can use. So they can provide the features in the car.

Q:My next question is about a computer device, 2 minutes similar for the personal coming device, the smart iot devices, the currency have the syncline that rely on the network service and the edge computing.

Because I have using the ioiot platforms called the me home or the media would which from the xiao mi, the smartphone managers, big company. It's used to be rely on the cloud so that last year their cloud server have and was down twice. And I I cannot control my smart home devices. So others like the homepage from apple is based on the edge commission, the local devices, such as the home port and ipad. You need to have the apple's brand, local devices. That's the only way you can control the smart devices. And even some sensors. The sony, I makes 500 cents in the senses. It can do some ai organisms, local in the image sensor and some of the sensors, because of the power consumptions, ai chips in your center, it's very power concern, power consumption.

So they want to separate sensor and the ice piece or the long logic circuit. What are the advantage and disadvantage of the inclined and edge computed team devices?

A:So to understand what is edge computing, edge computing is bringing the computing power close to the action, which is basically a small data centers. You can find them scattered all over the place. They study the market, they look at where there's a lot of traffic, a lot of subscribers, instead of having everything in a huge data centers sitting somewhere in a different state or a different country, do you have it something close to you?

And this gonna save latencies. Then the latency we will be very low. The bandwidth itself will be really preserved. Why? Because in order for me to turn off a light, I don't have to go through the internet, and that goes to another state thousand miles and then go to the lamp in my room. Its gonna be an area, gonna be a small data center here close where there's a lot of users for the devices. That's the beauty of this is. And this is why edge computing is a technology that is getting the attention of many companies. Intel, adele, microsoft, google, all of those companies. They wanna have this because this gonna save the bandwidth. Not everybody using the same bandwidth, the same  huge data center, at the same time, it's gonna free the time of the data center to do the calculations. By the end of the day, all the data that processed by the center of the edge computing device can be uploaded to the big data center to process it.

But at the right action, you need this. I'll give you an example, too. Tesla is one of the biggest companies that is pushing for edge computing because of their cars, because you have to make a decision in a split of second. And although for you to do this, you cannot wait until it goes to that data center and then come to you. But if there is an edge center or edge computing a data center closed by, the communication will be faster. So we can make that decision. Netflix is using this all the time based on how many people are watching a a certain episode of the movies.

So instead of having 20 big data centers in the country, you're gonna have hundreds of small centers that can process those order very quickly. And this is the difference between the two.

Q:The next is my last questions. It's about the matters compared of the a civilian market. The digital twins is relatively clear use cases in the industry field, because in the consumer market, every year is that this is the first year of the vr the first year of the mr such that but it's never go popular.

Or it didn't have civic use cases. Just answer one question, what's the vr work? What I can use in v rs? Almost everyone. I just said you can gaming, you can use the play a game and that, but I can use my screen to play a game. Right? Correct. It didn't tell give the consumers. What's the pacific use? The case only can provide it from the metaphors it didn't have for now. In the industrial field, we found more and more industry company are using data creep, the digital twins for their development that have developed a car to develop for the shape of cars. You did don't need to produce the product. The physical prototype, you can just design and build them in the metaphors.

So I agree with you. So what caused the difference between the metaphors from the matter was in the consumer market and the industrial market?

A:10 years, 2030. That's will be the time where you're gonna find the vr, ar, xr and metaverse is really becoming some kind of commodity where people can use it for doing their work, going to school, doing interviews, being in the office or not. But that's gonna take time. Because you need multiple things before it can get to the point.

For example, our interview here, it's 2D, two dimensional. If we have the metaverse, i'm gonna have an avatar that represent me and you gonna have an avatar represent you, we're gonna be sitting in front of each other. And we can talk as an avatar to each other and we could shake hands, I can show you something, you can show me something, I can design my office. I don't have to have a fake background, it's gonna be something gonna build. According to mark zackerelberg and every body who's working metaverse, it's about 10 years. So by 2030, they will see some kind of application for the metaverse. We already have some metaverse that you can go and visit. You have the one for facebook horizons. That's their own metaverse where people can go around and know each other and sit down and buy real estate and have some concerts there, but still at the early stages. So within 10 years, it's gonna show us all those applications, too early.

I mean some university have one or two classes of metaverse here in the united states. It's basically classes of biology or chemistry where they show them how to do things. But that is always the beginning. Nobody guarantee it's gonna succeed, but you will never know until you see it.And mara they spent about $13 billion so far, just to have the infrastructures and to have all the research on that.

Nobody said that's gonna be available tomorrow or 2025 or this is within the range of 10 years, we will see how's that gonna spread. Once you see the applications that people talking about working from home, the metaverse is a solution for that you can walk from home. And basically you can be in an office where people can see you, you can have your avatar. So you have to sit in the office and do your work. Cannot be sitting home and hide behind the camera. So a is a lot of endurance for it, but I don't see the application beside games for the consumers. As you said, for the industry, the digital twin is there, which is very helpful and save money and time for many of them, for the manufacturing and for the companies.

Until we take that one from manufacturing side or from the commercial and industrial, which is take us to the industrial internet of things. You take it from industrial internet of things to the internet of thing or the consumer internet of things.The money is in the industrial internet of things. This is where the money more than anything else, because the market and the internet of things you have limited number of customers, but they have tons of money. When you talk about the consumer internet of things is a lot of headaches with that because you're dealing with billions of consumers. And you have all different laws and regulation, privacy, all these kind of things you have to deal with.

责编: 武守哲
来源:爱集微 #集微访谈# #物联网# #元宇宙#
THE END

*此内容为集微网原创,著作权归集微网所有,爱集微,爱原创

关闭
加载

PDF 加载中...