In 2024, what's worth researching in the SLAM algorithms?

[Gatsby23]

The rise in the prominence of the SLAM problem can be attributed to the proliferation of applications in mobile robotics, particularly in environments where GNSS signals are unavailable, such as indoors, tunnels, caves, and similar settings. Consequently, research in SLAM predominantly focuses on localization. In 2024, the advent of affordable LiDAR sensors has substantially addressed the accuracy concerns in the localization aspect of SLAM. While occasional exceptions persist, SLAM algorithms demonstrate robust performance across most mobile robots, regardless of cost considerations. As a result, SLAM may now resemble a modular component within robotics research, like active mapping. This question: Is further research into SLAM algorithms warranted, or are we transitioning into an era of task-specific SLAM?

[Gatsby23]

SLAM问题成为热点的原因可以说是因于移动机器人应用的不断增多，特别是在GNSS信号不可用的环境中，如室内、隧道、洞穴和外星表面等。因此，SLAM的研究主要集中在定位方面。到了2024年，廉价LiDAR传感器的出现已大大解决了SLAM定位方面的精度问题。虽然偶尔还存在一些例外情况，但SLAM算法在大多数移动机器人在不考虑成本的情况下都表现出了强大的定位性能。因此，现在SLAM算法可能更像是机器人研究中的一个模块化组件，类似于主动建图。这引发了一个问题：SLAM算法本身是否还值得进一步研究？或者我们是不是正在过渡到一个特定任务的SLAM时代？通用性SLAM还存在吗？我们该如何给“通用性”下个定义？