A pendulum could have a tethered mass such as a passenger pod tethered from a fulcrum at any height halfway A and B. Now if you remove the tether and substitute a downramp, long fat section and then an upramp, you can cover long distances at high speed. Due to start and finish being at the same altitude, even acceleration and cruise are provided by gravity. Trading potential energy at A for velocity and back to arrive at B. Cruise is what takes most energy in long distance transport and evacuated tubes are great for that. Any losses during travel would be expressed as failing to reach destination, or loss of pentential energy (altitude). That latter can be solved with a simple elevator. The sorter the distance, the more likely to work. Even without tube evacuation. To combat velocity/time loss, pods could be weighed down. A 10 ton pods loses less speed due to air+roll drag than seen in a same size 5 ton pod. A long as low rolling resistance and safe passage can be guaranteed, a heavier pod is better or speed. As long as the elevator at destination is up for the task. I envision a network of MagLev or even rail, tube or open, with high placed station and optional elevators. All kinds of drive could be implemented on board pods to overcome losses, but they need to be provided all along the route. stations would be hubs with tracks going in all directions. Gravity provides >90 or even >99% of the energy needed to reach the end station. Passengers take the elevator to the start station at the top of a tower/skyscraper or hill. Cruise track could be on ground, underground, or deep underground to suit landscape, station layout, distance and desired travel time. Terminal cruise velocity for various altitude difference between start and cruise, assuming zero drag: 100m 44.29 m/s 159 kph 250m 70.02 m/s 252 kph 500m 99.03 m/s 356 kph 4000m (ocean floor passage) 280.1 m/s 1008 kph A 250 m tower releases a 5 ton pod to glide/roll down a given track from A to B, 50km, starting at 159 kph and arriving at 80 kph. Average speed is then still ~120kph or 25 minutes. The B staton elevator would do its job and the pod can go back. The same track and a 10 ton pod would achieve ~140kph average speed (21 minutes) and only require a stronger elevator to lift the pod over a smaller vertical distance back to starting height. Barely any extra energy losses for the significant time savings. Gravity could be complemented with a "launch" drive, external or on board. To allow for more shallow tunneling or a lower tower on either or both ends. To save time on long stretches, mid-journey drive could be used to keep cruise speed up and make the most of a tall B station available.