I would like to consult some of the senior researchers on employment in the field of Autonomous Driving and Planning. I still have no ideas of what's the job market like of Autonomous Driving and Planning in industry? Which subfield should I choose? And here are some of the comments or replies I got from Coffee Chat with others.

1. Which part of decision-making and planning (such as parking, trajectory prediction, behavior prediction, etc.) are used ? Because the main work of graduate students is decision-making and planning algorithms, I wonder if I can learn + decision-making and planning in the future. This way, the employment path will be slightly wider.

2. The fields of cooperative customers are in the fields of autonomous driving, non-road vehicles-autonomous driving, drones-agricultural plant protection machines, and low-altitude commercial aircraft. The keywords are robots and drones. In the fields of robots and drones, there is an algorithm that cannot be avoided in R&D positions. By consulting various professional materials and communicating with candidates, the algorithm positions in these two fields are summarized. This summary is more from the perspective of headhunters looking for people. 1. Perception algorithms Perception can be divided into two categories: one: perception positioning, map construction; the other: perception detection, tracking targets. The job descriptions issued by corporate HR are very similar.

3. For decision-making and planning algorithms, all companies now require the use of machine learning and deep learning large model methods to make the movement trajectory of the machine more flexible. Not like a fool who can't turn when encountering obstacles, nor like a drunk who walks unsteadily. Job descriptions are basically the same in all companies.

4. Compared with motion control algorithms, perception algorithms, and decision-making and planning algorithms, the company's products are robots. Regardless of whether the robot is humanoid or shaped, people who work on autonomous driving control algorithms in the automotive field are basically not considered.

5. Reinforcement learning algorithm The position of reinforcement learning algorithm is very special. The reason why I feel it is special may be that the information collected is not enough, because the companies that recruit for this position are all robot companies. In the field of autonomous driving, I have not seen any company recruiting positions that write "reinforcement learning algorithm" or "deep reinforcement learning algorithm".

6. At present, the interaction between autonomous driving in the pan-transportation field and the physical world is still relatively weak. Generally, it is enough to complete obstacle avoidance and passage. The embodied intelligence emphasizes the strong interaction between subjectivity and the physical world. It is more difficult. It is still in the early stage and the future situation is not clear. During my master's degree, I worked on trajectory prediction of autonomous driving vehicles, and now I am moving towards embodied intelligence.

7. Let's talk about robots first, especially foot-type and humanoid robots. The prospects are quite broad! With the continuous advancement of technology, they have great potential in industrial production, medical care, home services and other fields. Imagine that in the future, there will be a robot at home that can help you do housework and take care of the elderly, and the foot-type robot in the factory can work in various complex environments. How convenient it is! With the aging population and rising labor costs, the demand for such robots will increase. Not to mention autonomous driving, which is a major trend in future transportation. Once autonomous driving technology matures and is widely used, traffic accidents can be greatly reduced, traffic efficiency will be greatly improved, and people's travel methods will change fundamentally. The domestic traffic environment is complex, and the demand for autonomous driving technology is urgent, and there is a lot of room for development. Embodied intelligence, applied to the field of robotics, is simply opening the door to a new world! Allow robots to better understand and respond to human commands and interact with people more naturally. This is also an emerging and popular direction, with broad development prospects. Reinforcement learning provides powerful algorithm support for fields such as robotics and autonomous driving, allowing these technologies to be continuously optimized and improved. I think these directions are worth deepening. However, the specific choice depends on your interests and expertise. For example, if you are interested in human-computer interaction and service fields, embodied intelligence and humanoid robots may be more suitable for you; if you are passionate about the field of transportation, autonomous driving is a good choice. If you want to find foreign connections, you must first pay attention to international cutting-edge research results and academic trends. Participate in more international academic conferences and communicate and cooperate with foreign experts and scholars. You can also apply for foreign scientific research projects or study abroad to integrate into international cutting-edge research teams.

For more informtion, you can visit Robotic Search Engine to find latest research papers, AI Agent Marketplace and AI Agent Search Engine to find AI Agents related topics reviews and latest product information.