The antimatter was the lightest part of the shipment

The truck carried roughly 5.5 × 10²⁷ times more machine than antimatter.

Its cargo was 92 antiprotons, with a combined mass of about 1.54 × 10⁻²⁵ kilograms. Keeping that almost inconceivably small payload intact required an 850-kilogram apparatus: superconducting magnet, cryogenic trap, liquid helium, batteries, vacuum hardware, detectors and a frame sturdy enough to be lifted by crane.

That absurd ratio is the central fact of the experiment. Transporting antimatter is not primarily a problem of moving mass. It is a problem of maintaining an exceptionally controlled absence—no contact with ordinary matter, no loss of electrical confinement and no interruption while the laboratory becomes a vehicle.

In September 2026, the BASE collaboration reported that it had driven those 92 trapped antiprotons 7.5 kilometres through CERN’s Meyrin site without losing one. The trip took 24 minutes, included two laps in regular site traffic and reached 40.8 kilometres per hour. The full autonomous interval, including loading and reinstallation, lasted 2.72 hours (Nature paper).

A bottle whose walls cannot be touched

An antiproton meeting a proton can annihilate, so an ordinary container is useless. BASE-STEP instead uses a Penning trap: a magnetic field confines charged particles radially while an electrostatic potential confines them along the field axis.

The trap sits in the cold bore of a superconducting magnet. During laboratory operation the field can reach 1 tesla and the apparatus runs near 4.3 kelvin. For transport, the researchers lowered the field to 136 millitesla, reducing the energy stored in the magnet and therefore the consequences of a possible quench.

They then closed the valves separating the apparatus from its beamline, disconnected external power and cooling, and switched to an autonomous configuration. Batteries maintained the electrode voltages and monitoring electronics; liquid helium kept the magnet cold. A crane placed the complete frame on the truck.

The machine was not merely preserving the antiprotons. It was listening to them.

Trapped particles oscillate along the magnetic field and induce minute image currents in nearby electrodes. A superconducting resonant circuit, operating around 453 kilohertz, turns that collective motion into a dip in its electrical noise spectrum. For a small cloud, the width of the dip is proportional to the number of particles.

That gave the team a non-destructive passenger count throughout the journey.

The detector noticed the road

The ride did disturb the apparatus. During the first lap, the resonator frequency shifted by 122 hertz, probably because vibration slightly deformed a detection wire and changed the circuit’s capacitance. The antiprotons’ axial frequency also drifted slowly, which the researchers attributed to temperature variation in the precision voltage source.

These were instrument changes, not evidence of escaped particles. After the trap returned, the team retuned the antiprotons to the detector and compared long measurements taken before and after transport. The dip widths were 334.08(8) and 333.92(6) hertz. Losing one antiproton should have changed the width by 3.633(16) hertz—more than an order of magnitude larger than the observed difference (Nature paper).

That is the important observation: the experiment did not merely finish with some antimatter still present. Its calibrated signal excluded the loss of even one of the 92 particles.

The vacuum survived too. Across 33 days of storage, including four autonomous runs totalling 12.3 hours, the researchers observed no annihilation attributable to residual gas. One antiproton was lost on day 26 while the team was deliberately manipulating trap voltages, rather than during transport.

How much energy was actually aboard?

The word “antimatter” invites a science-fiction sense of scale. The arithmetic points in the opposite direction.

Using the recorded particle count, an antiproton rest-energy equivalent of 1.50327761802 × 10⁻¹⁰ joules, and an equal amount of ordinary matter for annihilation, the maximum mass-energy release is about 2.77 × 10⁻⁸ joules, or 27.7 nanojoules. That is enough energy to power a one-watt device for only 27.7 nanoseconds.

Here is the complete calculation:

from decimal import Decimal, getcontext

getcontext().prec = 30

particles = Decimal(92)
apparatus_kg = Decimal(850)
antiproton_kg = Decimal("1.67262192595e-27")
rest_energy_j = Decimal("1.50327761802e-10")

cargo_kg = particles * antiproton_kg
mass_ratio = apparatus_kg / cargo_kg
annihilation_j = Decimal(2) * particles * rest_energy_j

print("antiproton mass:", cargo_kg, "kg")
print("apparatus / cargo:", mass_ratio)
print("annihilation energy:", annihilation_j, "J")

The factor of two in the final line counts both the antiprotons and the equal mass of matter they would annihilate with. This calculation describes an energy ceiling, not what happened: the measured result was that the particles remained confined.

Moving the experiment away from its own noise

Why put such a delicate machine on a truck at all? At CERN’s Antimatter Factory, magnetic-field fluctuations from facility operations limit the precision of BASE’s comparisons between protons and antiprotons. Those comparisons test charge–parity–time symmetry, a foundational symmetry of modern particle physics.

BASE-STEP is intended to carry antiprotons to quieter laboratories where measurements can run with less ambient magnetic noise. The collaboration expects that strategy could eventually improve its symmetry tests by at least a factor of 100. That improvement is a goal, not a result demonstrated by this drive.

What the drive established is narrower and wonderfully concrete: a cryogenic Penning trap can leave its beamline, hang from a crane, ride twice around a working research site and return with its count of antiprotons unchanged. The antimatter barely contributed to the shipment’s mass. Nearly everything aboard existed to preserve the tiny patch of empty space around it.