HUNTER: Highlight the Unexpected with Navigation Through Extreme Regions

政府间气候变化专门委员会(IPCC)在《2019年气候变化中的海洋和冰冻圈特别报告》中指出，适应气候变化意味着应对极端气象事件增加的不确定性。此外，有证据表明，北极海冰的减少导致北半球的大雪事件增加。最后，改进恶劣天气条件下自主导航的感知算法目前是一个开放的挑战。支持创建此类算法的实验需要安全地进入降雪概率较高的地区，需要专门的机器人设备，以及在自动导航方面的大量科学知识，这在单一实验室内是很难找到的。多伦多大学在状态估计理论和自动车辆基于摄像头的导航解决方案方面拥有成熟的专业知识。在巴尔福特教授的领导下，在过去的二十年里，通过将复杂的机器人现场部署和坚实的理论背景相结合，实现了许多创新应用和关键科学出版物。近年来，该实验室开始研究使用雷达代替相机进行定位。雷达的主要优势是能够在大雪中感知障碍物，但代价是测量的不确定性更大。三年前，随着北方机器人实验室的成立，拉瓦尔大学在冰冻圈自动导航方面拥有新兴的专业知识。通过战略基础设施投资和与工业合作伙伴的长期合作，由波梅洛和吉古埃尔教授领导的实验室能够获得能够在深雪覆盖的情况下导航的大型自动地面车辆。通过关键出版物和国际声誉，弗朗索瓦·波梅罗在基于激光雷达的导航算法方面建立了专业知识，这对共同申请者的专业知识起到了很好的补充作用。最后，这所大学拥有一片250平方公里的研究森林，每年降雪量为6米。这个装置可以容纳研究人员很多天，是这个合作研究项目的核心。通过结合这个跨省团队的优势，这个研究项目旨在验证冬季条件下越野导航的自主导航解决方案。通过将坚实的理论基础与基于大规模现场部署的实验证据相结合，将有可能发现针对冬季条件产生的巨大不确定性的稳健算法。这一计划的结果将提高这两个机构在恶劣天气条件下的自主权的国际知名度。此外，这些专业知识将支持政策制定者在我们的道路上采用自动驾驶汽车，同时增加自动驾驶汽车对北方社区的可及性。

In the 2019 Special Report on the Ocean and Cryosphere in a Changing Climate, the Intergovernmental Panel on Climate Change (IPCC) states that climate change adaptation implies dealing with the uncertainty of increasing extreme meteorological events. Moreover, there is evidence that the decline of the Arctic sea ice led to an increase in heavy snowfall events in the Northern Hemisphere. Finally, improving perception algorithms for autonomous navigation in poor weather conditions is currently an open challenge. Experiments supporting the creation of such algorithms require safe access to an area with a high probability of snowfall, specialized robotic equipment, and large scientific expertise on autonomous navigation that is difficult to find within a single laboratory.The University of Toronto has well-established expertise in state-estimation theory and camera-based navigation solutions for autonomous vehicles. Led by Professor Barfoot, many innovative applications and key scientific publications were achieved in the last two decades by combining complex field deployments of robots and solid theoretical background. In recent years, the laboratory started to investigate the use of radar as a substitute for cameras for localization. Radar has the key advantage of being able to perceive obstacles through heavy snowfall, at the cost of more uncertainty of measurements.Laval University has emerging expertise in autonomous navigation in the cryosphere with the creation of the Northern Robotics Laboratory three years ago. Through strategic infrastructure investments and long-term collaborations with industrial partners, the laboratory, led by professors Pomerleau and Giguère, has access to a large-scale autonomous ground vehicle capable of navigation over deep snow covers. Through key publications and international accolades, François Pomerleau has established expertise in lidar-based navigation algorithms, which complements the co-applicant expertise well. Finally, the university owns a research forest of 250 square kilometres over which six meters of snow are falling every year. This installation is equipped to accommodate researchers for many days and is the centerpiece of this collaborative research program.By combining the strengths of this interprovincial team, this research program aims at validating autonomous navigation solutions for off-road navigation in winter conditions. By combining strong theoretical foundations with experimental evidence based on large-scale field deployments, it will be possible to discover robust algorithms against the large uncertainty that winter conditions produce. Outcomes of this program will augment the international visibility of both institutions on autonomy in harsh weather conditions. Moreover, the expertise will support policymakers for the adoption of autonomous cars on our roads, while augmenting the accessibility of autonomous vehicles to Northern communities.