After the last scientific collisions in June 2026, the LHC will be shut down for a major upgrade. When it returns in 2030 as the HiLumi LHC, it will produce many more collisions and enable a precise search for new physics.
The Large Hadron Collider, LHC, has entered the longest and most complex shutdown in its history. The collisions designed to collect scientific data ended on June 14, 2026, and after two weeks of high-intensity proton beam experiments, the accelerator will be completely shut down on June 29. Over the next four years, key parts of the accelerator and detectors will be replaced, with a view to resuming operations in the mid-2030s under the name The Large Hadron Collider, a high-luminosity – High-Luminosity LHC, or for short HiLumi LHC. ([CERN][1])
The LHC is a 27-kilometer-long ring, installed about a hundred meters below the border between France and Switzerland. Superconducting magnets direct two beams of protons moving in opposite directions, causing them to collide inside the large detectors. ATLAS, CMS, ALICE and LHCb. The goal is not to “smash” matter for the sake of destroying it, but to momentarily create conditions of high energy density, in which elementary particles and rare phenomena can be produced and studied. ([CERN][2])
Not much more energy – but a lot more collisions
Unlike previous upgrades, the main change this time is not a dramatic increase in collision energy. The goal of the project is to increase the The clarity – Luminosity – A measure related to the number of collisions that occur over a given time. The more collisions occur, the greater the chance that one of the rare events that physicists are looking for will actually appear in the data.
CERN plans to increase the combined luminosity of the accelerator tenfold compared to the original design value. Each encounter between two proton beams is expected to result in between 140 and 200 overlapping collisions, compared to about 60 today. During the years of operation of the HiLumi LHC, the experiments are expected to accumulate up to 4,000 inverse femtobarren of data – several times the total information collected in the first three run periods. ([CERN][3])
To this end, components along about 1.2 kilometers of the ring will be completely replaced. New magnet systems will produce denser and more focused proton beams near the collision points. At the same time, the ATLAS and CMS detectors will undergo extensive upgrades, including the installation of new silicon detectors, more precise timing systems, and electronics capable of withstanding radiation and high data rates. ([CERN][3])
Hundreds of millions of Higgs bosons
The most well-known result of the LHC is the discovery Higgs boson In 2012, the ATLAS and CMS experiments confirmed the Higgs mechanism, which explains how elementary particles acquire mass through their interaction with the Higgs field. However, many questions about the particle remain unanswered, including the strength of its interaction with itself and the possibility that it is linked to particles not included in the Standard Model. ([CERN][4])
Since its inception, the LHC is estimated to have produced about 55 million Higgs bosons. CERN estimates that the high-luminosity version will produce about 380 million Higgs bosons over its lifetime. The large number will allow for more precise measurements of rare decays and to test whether the particle's properties are fully consistent with the predictions of the Standard Model. Even a small deviation could hint at new physics. ([CERN][3])
One intriguing goal is to clearly detect the production of two Higgs bosons in the same collision. The process is supposed to occur according to the Standard Model, but it is so rare that it has not yet been observed with a level of certainty that allows us to declare its discovery. Studying Higgs pairs may allow us to directly measure the self-interaction of the Higgs field – a fact that also relates to the evolution of the universe shortly after the Big Bang. According to CMS estimates, the first evidence of the process may appear after about half of the HiLumi LHC data have been collected. ([CMS][5])
Will the upgrade reveal dark matter?
Many headlines portray the upgrade as a journey of discovery dark matter, but it is not certain that dark matter particles are within the range that the LHC can detect. If such particles are created in a collision, they will probably not react with the detectors and will pass through them without leaving a trace. Physicists will therefore look for an imbalance in the energy and momentum of the particles measured – a kind of “missing signature” that might indicate that an invisible particle has escaped the detector. ([CERN][6])
More data will allow us to examine rarer processes and place tighter limits on models of dark matter, supersymmetry, extra dimensions, and long-lived particles. But even if no new particle is discovered, a negative result will rule out entire families of models and narrow the field of possibilities. Dark energy, on the other hand, is not a direct target of the LHC collision experiments.
Artificial intelligence will help you choose what to keep
The detectors cannot record every collision that occurs in the accelerator. Systems called “triggers” analyze the data in real time and select the events that may contain interesting signatures. In the HiLumi LHC era, machine learning methods will become increasingly important, helping to distinguish within a fraction of a second between routine collisions and unusual events that should be recorded for further analysis. ([ATLAS Experiment at CERN][7])
Artificial intelligence will not replace physicists and will not decide what a discovery is. It will serve as a tool for filtering and reconstructing data under conditions where hundreds of collisions overlap. Deciding whether a particular phenomenon does indeed indicate a new particle will continue to require statistical analysis, repeated tests, and comparisons between several independent experiments.
The decommissioned accelerator has already provided each of the two major experiments, ATLAS and CMS, with some 54 quadrillion proton collisions and has led to some 4,500 peer-reviewed scientific papers. The data already collected will continue to be studied during the upgrade. When the accelerator comes back online in 2030, it will not be a completely new machine – but a much more sensitive, faster and more precise version of the facility that revolutionized particle physics. ([CERN][1])
Questions and Answers
Why is the LHC being shut down for four years?
To replace systems along about 1.2 kilometers of the accelerator, install new magnets, and upgrade the detectors, cooling systems, electronics, and data processing.
Will HiLumi LHC be a higher-energy accelerator?
The main change will be in the number of collisions, not the maximum energy. Dense, more focused beams will produce much more data in the same period of operation.
Is it likely to discover dark matter?
Indirect clues may appear, such as missing momentum or new intermediary particles, but there is no guarantee that dark matter can be produced at the energies of the LHC.
For the original publication: Opening the original publication
More on the subject on the science website
- Not just predicting words: World models that try to teach artificial intelligence to predict the results of its actions
- The first nuclear clocks have begun ticking – opening a window to new physics
- Artificial intelligence helps decipher how heavy elements in the universe are formed
- More articles, fewer ideas? The danger of using artificial intelligence in science
- Black holes from before the Big Bang? Cosmological model suggests ancient location “fossils”
