Conventional elevators are, at their core, 19th-century technology: a box on a rope. Steel cables loop over a sheave, a counterweight balances the load, and the whole system is constrained to a single vertical shaft. The tallest buildings already push this design to its breaking point — the Jeddah Tower's elevators require carbon-fiber ropes because steel cables would collapse under their own weight past about 500 meters. So what if we ditched cables entirely and drove elevator cars with linear induction motors, like a vertical maglev?

This isn't purely hypothetical. ThyssenKrupp's MULTI system, first demonstrated in 2017, does exactly this. But let's dig into the physics of scaling it up and ask: what does it actually take to levitate and launch a fully loaded elevator car straight up a kilometer-tall shaft?

The force budget. A typical elevator car masses about 2,500 kg empty. Add 20 passengers at 80 kg each: that's 4,100 kg total. To hold it stationary against gravity requires a continuous thrust of:

F = mg = 4,100 kg × 9.81 m/s² = 40,221 N ≈ 40 kN

That's just to hover. To accelerate at a comfortable 1.2 m/s² (typical high-rise elevator), we add:

F_accel = 4,100 × 1.2 = 4,920 N

Total peak thrust: about 45 kN. For context, a Tesla Model S rear motor produces around 50 kN — so we're in the ballpark of a single large electric motor, which is reassuring.

The power question. At a cruising speed of 10 m/s (comparable to fast elevators today), the sustained power against gravity alone is:

P = Fv = 40,221 N × 10 m/s = 402 kW ≈ 539 horsepower

That's per car, continuously, with no counterweight giving you a free ride. A conventional counterweighted elevator at the same speed draws only the power to overcome the imbalance — maybe 30-50 kW. Removing the counterweight multiplies energy consumption by roughly 8-10×.

So why bother? Because cables constrain you to one car per shaft, moving only vertically. A linear-motor car on guide rails can switch between vertical and horizontal tracks. Suddenly shafts become a network. Multiple cars share the same loop, like a paternoster on demand. In a supertall building, you might run 8 cars in a shaft that previously held one, slashing wait times from 40 seconds to under 15. The shaft footprint — which eats 25-30% of a skyscraper's core — could shrink dramatically.

Recovering that energy penalty. Regenerative braking on descending cars is straightforward: a loaded car going down generates almost as much power as an ascending one consumes. In a busy building with balanced traffic, net energy approaches that of a counterweighted system. The ugly case is off-peak hours when cars travel empty in one direction — but even then, supercapacitor banks in the basement can buffer regenerated energy for the next ascending trip, recovering 70-85% of the gravitational potential.

The thermal wall. Linear induction motors are less efficient than rotary ones — typically 75-85% versus 95%. At 402 kW input, the motor's stator plates lining the shaft dump 60-100 kW of waste heat per car into a confined concrete column. With 6 cars running, that's 600 kW in an enclosed shaft. You essentially need to air-condition your elevator shaft, which is an engineering sentence nobody wanted to write.

Still, the math closes. The energy penalty is real but manageable. The throughput gains are transformative. And for buildings over 500 meters, it's not just better — it's the only architecture that works without exotic rope materials.