State estimation for robots navigating in GPS-denied and\nperceptually-degraded environments, such as underground tunnels, mines and\nplanetary subsurface voids, remains challenging in robotics. Towards this goal,\nwe present LION (Lidar-Inertial Observability-Aware Navigator), which is part\nof the state estimation framework developed by the team CoSTAR for the DARPA\nSubterranean Challenge, where the team achieved second and first places in the\nTunnel and Urban circuits in August 2019 and February 2020, respectively. LION\nprovides high-rate odometry estimates by fusing high-frequency inertial data\nfrom an IMU and low-rate relative pose estimates from a lidar via a fixed-lag\nsliding window smoother. LION does not require knowledge of relative\npositioning between lidar and IMU, as the extrinsic calibration is estimated\nonline. In addition, LION is able to self-assess its performance using an\nobservability metric that evaluates whether the pose estimate is geometrically\nill-constrained. Odometry and confidence estimates are used by HeRO, a\nsupervisory algorithm that provides robust estimates by switching between\ndifferent odometry sources. In this paper we benchmark the performance of LION\nin perceptually-degraded subterranean environments, demonstrating its high\ntechnology readiness level for deployment in the field.\n