11月27日下午，在2021亚洲青年领袖论坛“科技创新与智慧城市”主题论坛平行对话中，小马智行副总裁莫璐怡就“无人驾驶技术创新迭代的亮点与痛点”这一话题发表了自己的观点。

摘录如下：

很高兴今天来到第一届亚洲青年领袖论坛，跟大家做一个分享。我先做个简单的自我介绍：我叫莫璐怡，来自小马智行，我是小马智行的副总裁，也是小马智行Robotaxi业务的负责人。欢迎大家来到广州南沙。广州南沙是小马智行中国的总部所在地，我自己本身也是土生土长的广州人。所以欢迎各位来到我的故乡，来到我们小马智行中国的总部所在地。如果各位接下来有空余的时间，可以拿出你们的手机，尝试打一下我们小马智行在广州的Robotaxi去体验一下自动驾驶的新技术。

关于智慧城市，首先，我想先谈谈我认为的智慧城市它的目标是什么。我觉得智慧城市是为人民未来的生活去提供一个安全、可靠、高效的生活服务的。所以，所有的智慧城市的建设都是围绕着这个话题。交通必然是智慧城市建设当中一个不可或缺的环节，特别是人的出行和物的运输又是交通领域的两大主题。那么接下来，我想通过小马智行在自动驾驶领域过去5年的实践来分享一下我们在智慧交通建设、在人的出行和物的运输方面，所探索的新场景和所使用的新技术以及我们希望给未来人们提供的新体验。

小马智行是2016年底，也就是5年前，在美国硅谷成立的。成立的4个月之后就成立了中国研发中心。在2017年底，我们把中国总部落在了南沙。2018年2月，我们搭载了小马智行全新一代的自动驾驶软硬件系统的广汽GE3就在广州南沙正式驶上了公开道路，开始了常态化的自动驾驶测试，这是中国第一次在公开道路上面看到了自动驾驶的车辆。从2018年2月至今这四年当中，不管是行业本身，还是我们小马智行本身，也是经历了非常多的变化。从当初只是在南沙区政府到市民广场的一段大概几百米的道路，到今天我们已经可以在广州南沙的全域超过300平方公里的核心城区，以及在北京亦庄全域和上海的嘉定，都推出了来自我们小马智行的自动驾驶出租车的服务。现在非常方便，大家每天可以用我们的APP，就像网约车一样，就可以在刚刚所说的广州、北京和上海三个地方去乘坐我们自动驾驶出租车，去进行日常的通行。不管是在大雨天，还是在高峰期繁忙的城市道路当中，您都会发现来自小马智行的机器人司机，可以带你去任何一个你想去的地方。这是一种全新的体验，它带来的是一种安全、舒适又高效的出行。

除此之外，在货运方面，我们也有很多的技术应用在自动驾驶的重卡车上面，去代替人类司机。因为我们发现在货运的行业，人力短缺是非常重要的一个问题，特别是在现在人口老龄化加速的情况之下。那么，今天的机器人司机就可以很好地解决这样的困难。同时，它还能够极大地解决并提高整个出行或运输过程当中的安全性。因为我们发现90%的交通事故其实都是人为的事故，所以通过这样的自动驾驶的新技术，我们在出行的领域和运输的领域，都可以去感受到一个来自于未来的全新体验。

我觉得说到痛，或者其实应该是说我们要面对的挑战。在整个无人驾驶的发展过程当中，我们的挑战是来自两个方面的。第一方面是来自于企业自身的技术发展；第二方面是来自于把技术应用到社会当中所引起的，包括像法律、法规等等这样的一些社会问题所带来的挑战。那么，我先讲讲第一个挑战，即关于技术上的挑战。在无人驾驶的发展历程当中，我们认为两个话题是非常重要的，必须要先做到，我们才能谈到无人驾驶的规模化的应用。说起无人车，第一个就是无人。要做到技术足够的安全，才可以去做到这样的无人的水平，这里面很多技术的挑战需要去面对，需要去解决。因为道路上面有非常多突发的情况。那么，我们的整个系统面对各式各样的突发情况，它都需要去灵活应对，需要去保证安全，同时又要去提升舒适度和通行的效率，这里面是有非常多的挑战的。

在今年6月份的时候，我们的两位创始人在南沙的路上坐上了我们的无人车。我们为什么要在那个时间点开始去做这个事情？其实，我们会去不断地分析我们整个系统的安全性，我们认为无人驾驶的目标，其实是提高交通的安全性，它要做到比人更安全。所以，像今天在南沙的区域当中，我们在一系列非常详尽的评判指标下如何去分析系统能力很重要。我们过去其实也在一直提升这方面的能力。在这样的一个评价指标之下我们发现，今天我们的系统其实已经可以做到比人还要更安全，但是它离真正大规模的应用还是有一定的距离的。所以，其中很重要的挑战是我们如何去不断地提升，不断地迭代，去面对更加复杂的场景和更加多变的道路情况。

第二个痛点，就是我们要使得自动驾驶可以大规模的落地。我们要解决一个问题，就是规模化的问题，我们要去做量产。所以，我们今天跟包括广汽在内的很多国内外的主机厂都有非常深度的合作，其实就是为了解决量产这样的问题。因为自动驾驶系统和无人驾驶系统，它必须要跟车辆非常有机地结合在一块。就像人的大脑跟身体一样，你必须要协调，你才能跑得快，自动驾驶也一样。我们需要跟车辆有很好的协调，然后我们把这样的技术做到可以规模量产的一个阶段，这就是我们在技术上面的第二个痛点。无人化和规模化的实现，才是我们无人驾驶可以真正走向大规模应用的前提。我觉得这些更多是企业自身需要去不断解决的一些痛点，一些必须迎接的挑战。

另外就是我们将这样的技术真正应用在社会当中的时候。像刚刚各位嘉宾也都提到了相关的话题，这个时候需要对法律法规有很多的创新。这一点上面，我也有两个观点。一是我们当然非常希望整个政策的开拓创新，可以对新技术有足够的包容，去鼓励创新创业的实践，但另一方面，政策本身也需要不断地完善，需要提高门槛，需要设定标准，设立行业的指标，这样才能保证这一个行业可以持续不断地健康发展。在这一点上，我们与同行业上下游很多的合作伙伴和政府的相关部门做了很多的探讨，比如立法部门，不断地去针对相关话题，不管是政策的话题还是标准的话题进行反复探讨。

在刚刚过去的一周，我们在北京获取了中国首个Robotaxi的商业示范的许可。这个许可新增加了非常多的监管的要求，需要技术达到一定水平，需要在安全上，不仅是车辆系统安全，也包括信息安全、数据安全等方面都要满足新的要求才能获得。这也是我们跟相关监管部门在探讨的过程当中一起推动出来的。我们相信有一个高的门槛和一个好的行业标准是有利于整个行业未来发展的。这是我们放眼未来的五年、十年之后，当整个无人驾驶技术进一步成熟之后，当它大规模地应用在我们生活的方方面面的时候非常需要去解决的问题。

最后我们希望通过小马智行联合其他的相关企业，一起为这个行业的发展发挥更大的作用。


On 27 November, Mo Luyi, Vice President of Pony.ai, shared her opinions on the highlights and challenges of innovation in self-driving technologies at the Technology Innovation and Smart City Forum of the Asia Youth Leaders Forum 2021.

The full speech is as below.

I'm very happy to attend the Asia Youth Leaders Forum today and share my views with you. Allow me to make a brief self-introduction first. My name is Mo Luyi. I'm the Vice President of Pony.ai and the head of the Robotaxi business of Pony.ai. I want to welcome you all to Nansha. The China headquarters of Pony.ai is in Nansha. And I'm a native of Guangzhou. Welcome to my hometown as well as the China headquarters of Pony.ai. If you have time, you can use your mobile phone to try the Robotaxi service of Pony.ai in Guangzhou and experience the new technology of autonomous driving.

Regarding smart cities, first of all, I want to talk about what the goal of smart cities is from my understanding. I think a smart city is one that provides safe, reliable, and efficient lifestyle services for future life. And all smart city development should be based on this. Transportation is an indispensable part in the building of smart cities. The mobility of people and the transportation of goods are two main parts in the field of transportation. Next, I would like to share what we have been doing in the past five years in autonomous driving, including some new scenarios we've explored, new technologies we've used, and new experiences we want to give people in the future during our development of smart transportation and the transportation of people and goods. 

Allow me to introduce you to Pony.ai. Our company was established in Silicon Valley in the United States five years ago, at the end of 2016. Only 4 months after its establishment, our China R&D Center was established. At the end of 2017, we set up our China headquarters in Nansha. In February 2018, the GAC GE3 vehicles equipped with our latest self-driving software and hardware system at the time hit the road in Nansha and began regular autonomous driving tests. That was the first time China saw an autonomous vehicle on public streets.

It has been four years since February 2018. During these four years, the industry and Pony.ai have gone through a lot of changes. At that time, we test-drove the cars on a path of several hundred meters in Nansha, between the district government and Citizen Square. Today, we have launched the self-driving taxi service in Nansha, covering a core urban area of more than 300 square kilometers in the district, as well as in Beijing Economic-Technological Development Area and Jiading District of Shanghai. Now it is very convenient to use our service. It's like a ride-hailing service. You can take our self-driving taxis in the aforementioned three places in Guangzhou, Beijing, and Shanghai for daily travel needs. Whether it's on a rainy day or on busy city streets during rush hour, you can see robot drivers from Pony.ai who can take you wherever you want. It's a whole new experience: a safe, comfortable, and efficient travel experience. In addition, in terms of freight, we apply self-driving technology to heavy trucks to replace human drivers. We have found that in the freight industry, manpower shortage is a very serious problem, especially with the accelerated aging of the population. But with our robot drivers, this problem can be solved very easily. At the same time, it can greatly improve the safety of the entire travel or transportation process. We have found that 90% of traffic accidents are actually caused by human drivers. Through new technologies such as autonomous driving, we can experience something from the future in the field of transportation. 

When it comes to the challenges we have been facing in the development of autonomous driving, there are two aspects. The first aspect is the challenges of technological development in the company. And the second is the challenges from applying the technology to real life, including laws, regulations, and other social problems. Now I'll talk about the first aspect: technological challenges. In the development process of autonomous driving, two issues are very important, which must be settled before large-scale application can be carried out.

The first issue is autonomy. For a vehicle to be safe enough for autonomous driving, many technological challenges need to be solved, because there are many unexpected situations on the road. In the face of various emergencies, our system needs to respond flexibly and ensure the safety of passengers. At the same time, it is necessary to improve the comfort and efficiency of the rides. This means a lot of challenges. In June this year, our two founders got on one of our autonomous vehicles on the road in Nansha. Why would we start doing this at that point in time? In fact, we have been analyzing the security of the entire system. We think that the goal of autonomous vehicles is to improve traffic safety. So, they must firstly be safer to ride than with human drivers. Like in Nansha, how to analyze the system’s capability against a series of detailed evaluation indicators is very important. We have actually been improving our capabilities in this area. Against such evaluation indicators, we have found that our system can actually be better than humans in safe driving. However, there is still distance from large-scale application. One of the most important challenges is how to continuously improve the system to face more complex scenarios and ever-changing road conditions. 
Then, what is the second issue? If we want to apply autonomous driving on a large scale, we have to solve the problem of scaling or mass production. This is why we have deep cooperation with many domestic and foreign automobile manufacturers including GAC, to solve the problem of mass production. Autonomous driving system must be organically integrated with the vehicle, just like the brain and body of a human being, which must be coordinated in order to run fast. Our system needs to be well coordinated with the vehicle before we can apply this technology on a large scale. This is our second issue in the technology aspect. The realization of autonomy and large-scale application is the prerequisite for the wide deployment of autonomous driving. I think these are the challenges that enterprises themselves need to overcome.

In addition, when we apply such technology to real life, there needs to be innovation in laws and regulations. In this regard, I also have two points of view. On one hand, of course, we very much hope that the policy is pioneering, innovative, and sufficiently inclusive of new technologies and encourages the practice of innovation and entrepreneurship. On the other hand, policies also need to be continuously improved. Policies need to raise the threshold and set standards and industry indicators. Only in this way can we ensure sustainable and healthy development of this industry. So, regarding these issues, be it policies or industry standards, we have conducted many in-depth discussions with upstream and downstream partners in the industry as well as relevant government agencies such as legislative departments. This past week, we obtained a license for China’s first commercial demonstration of Robotaxi service in Beijing. To obtain this license, we need to meet many new regulatory requirements, such as requirements for technology and security, including vehicle system security, information security, and data security. This is also an outcome of our discussion with relevant regulatory authorities. We believe that a high threshold and good industry standard can benefit the future development of the entire industry. In particular, in the coming five to ten years when autonomous driving further matures and is applied in every aspect of our life, this problem will become one that must be solved. 
In the end, we hope that through our efforts at Pony.ai and cooperation with other relevant companies, we can make greater contribution to the development of this industry.