"Failsafe" for VLANs and gateways was the traditional approach to trigger a failover in case the active unit might become isolated from the infrastructure. Be aware, that the same failover condition might force both units to get into standby mode (resulting in a standby/standby).
The concept of "HA-groups" was introduced to replace the "failsafe" mode. With "HA-groups" a sub-second failover might be triggered because you are able to monitor the number of available links in a trunk (aka link aggregation with LACP in fast mode). The scoring allows a pretty granular selection of relevant resources to trigger a failover. One unit will always be active.