Throughout every day, particles produced by cosmic rays are passing through your body, the walls around you, and the ground beneath your feet. Most people will never notice them. A small detector called CosmicWatch turns those invisible arrivals into flashes, counts, and datasets that high school and university students can study with the same basic principles used by professional researchers.
CosmicWatch began in 2017 as a project to develop a compact, affordable muon counter that could travel to classrooms, balloons, and underground laboratories. The latest version, the CosmicWatch Desktop Muon Detector v3X, still costs less than $100 in components, yet adds faster electronics, environmental sensors, and the ability to connect multiple units. It is not a miniature particle accelerator or a complete observatory, but it provides a remarkably capable introduction to experimental physics.
The CosmicWatch v3X technical paper describes a device weighing about 110 grams, drawing 0.5 watts and using a 5-by-5-by-1-centimeter plastic scintillator to register passing particles.
What Is CosmicWatch?
The partly open-source CosmicWatch desktop particle detector is built around four main components:
- A slab of plastic scintillator, a material that absorbs energy from a passing charged particle and briefly re-emits it as light.
- A silicon photomultiplier, or SiPM, a tiny light sensor that converts flashes into electrical pulses.
- Custom electronics that amplify, timestamp, and measure each pulse.
- A small microcontroller with an OLED screen, microSD card, USB connection, and environmental sensors.
The unit flashes when its electronics register an event above a selected threshold. A single detector counts ionizing radiation and other particles, but it cannot prove that every event is a cosmic-ray muon. Random environmental radiation and other charged particles can trigger the sensor.
That distinction matters. CosmicWatch does not directly detect the original high-energy cosmic ray that began far beyond the Solar System. It usually records secondary particles created when that energy strikes molecules high in Earth’s atmosphere.
How Do Cosmic Rays Produce the Muons It Detects?
The term cosmic ray is a traditional name for a mixture of highly energetic particles, especially protons and atomic nuclei. They are not all identical, and many are not electrically neutral. Their origins can include the Sun, supernovae, gamma-ray bursts, active galaxies, and other energetic events, although scientists cannot identify the exact source of every individual arrival.
When a primary cosmic ray strikes an atmospheric nucleus, it creates a shower of secondary particles. Those particles collide again and produce many more. Muons form during this cascade and can travel at speeds close to that of light.
Many muons decay before reaching sea level, but those produced high in the atmosphere have enough energy to travel great distances. Some penetrate buildings, hills, and rock. Their abundance at ground level gives researchers a convenient messenger from an event they cannot observe directly.
Muons also helped provide one of the earliest experimental confirmations of Albert Einstein’s theory of special relativity. A moving muon should age more slowly according to its own clock, allowing some to survive a longer journey through the atmosphere than basic physics would predict.
How Does a Muon Trigger a Flash?
Inside the v3X detector is a 5-by-5-by-1-centimeter slab of polystyrene-based plastic containing fluorescent additives. When a charged particle passes through it, energy deposited along the particle’s track excites the material.
The scintillator then releases a burst of violet light. Reflective wrapping directs more of that light toward the 6-by-6-millimeter active area of the SiPM. The sensor converts individual photons into a brief electrical pulse. Custom analog electronics amplify the pulse and hold its peak long enough for the microcontroller to measure it.
The firmware records the timestamp, signal amplitude, live-time correction, temperature, pressure, and motion data. It can show a running count on the display, save detailed events to a microSD card, or stream the information through USB.
The v3X paper reports a maximum logged rate of 700 events per second, with a configurable mode capable of writing faster to the card when serial output is disabled. These are instrument capabilities, not guaranteed counts of muons arriving at the surface.
Why Students Need Two Detectors to Select Muons
One CosmicWatch can begin teaching radiation detection, electronics, programming, and statistics. The wider CosmicWatch program also supports lessons on beta and gamma radiation, detector design, printed circuit boards, data visualization, and basic statistics. A two-detector setup can do more: it looks for coincidence, meaning both units register signals close enough together in time to plausibly come from the same particle.
Random gamma and beta radiation is unlikely to trigger two separated units in the required microsecond window at precisely the right moment. A stacked or angled two-detector arrangement therefore rejects more background and provides a clearer selection of through-going cosmic-ray muons.
Depending on their separation and orientation, two units can also provide a rough indication of direction. The v3X paper shows a pair can reproduce a sea-level angular pattern close to the expected squared-cosine dependence of sea-level muon intensity. That is an introductory version of the same coincidence principle used in larger research instruments.
Unlike the miniature particle accelerator built by researchers at Stanford and SLAC, CosmicWatch does not accelerate particles. It passively records particles that have already reached it.
What Can Students Measure With CosmicWatch?
The best classroom activities begin with a question that can be tested using a count and a control. Examples include:
- Does the recorded event rate change between the ground floor and a rooftop?
- Does adding several centimeters of dense shielding change the accepted count rate?
- How does the response differ when the detectors are horizontal, vertical, or angled?
- Does barometric pressure or temperature correlate with the recorded rate?
- Can a pair of detectors measure how long a muon takes to cross the gap between them?
The right response to each question depends on data quality, live time, background rejection, detector calibration, and controlled conditions. A blinking light is an engaging demonstration, but a defensible experiment requires a clearly defined variable, repeated runs, uncertainty, and records of the detector settings.
The v3X data files are tab-delimited text, which makes the measurements accessible to spreadsheet and Python users. The latest design also logs cumulative dead time. Researchers should divide events by live time, not simply the total clock time, when calculating rates. A [QuarkNet and University of Notre Dame classroom study](https://pos.sissa.it/444/1630/pdf) found that students should generally collect at least 20 coincidence counts for a reasonably stable classroom rate and about 50 for greater reliability. Those recommendations apply to the study’s detectors and setup, not every CosmicWatch experiment.
From Classrooms to Balloons and Dark-Matter Experiments
CosmicWatch began with Spencer Axani, now a University of Delaware physics professor, while he was a graduate student at the Massachusetts Institute of Technology. His first aim was to create a small, low-power muon counter for the IceCube Neutrino Observatory in Antarctica. Portable, affordable hardware made the design useful far beyond its original purpose.
According to the University of Delaware, thousands of devices have been built since the project began. Units have been used in introductory and advanced laboratory courses at MIT, Cornell, and the University of Delaware, as well as in international school research and dark-matter instrumentation. The University of Delaware also reports that the v3X is being used in its NuDot experiment and in the Coherent CAPTAIN-Mills dark-matter detector at Los Alamos National Laboratory.
One University of Delaware balloon payload used paired detectors during a flight to 100,000 feet, about 30.5 kilometers. Its pressure data helped convert the flight profile into altitude, allowing researchers to observe the change in ionizing radiation with height. After the balloon burst, researchers recovered the detector miles from the launch site with its measurements intact.
Students do not need a giant accelerator to begin meaningful physics. The contrast with the Future Circular Collider proposed at CERN shows how a simple passive counter can open an entirely different experimental path.
Could CosmicWatch Become a Global Citizen-Science Network?
One of the project’s most ambitious ideas is to connect thousands of inexpensive detectors around the world. Participants would record local event rates and send observations to a central database, creating a larger map of cosmic radiation and other environmental sources of ionizing radiation.
Axani envisions using such a network to reconstruct extended atmospheric particle showers by combining measurements from widely separated locations. A future version of the detector is also being developed for rocket and spacecraft instrumentation.
Calibrating distributed measurements will be the difficult part. Instrument versions, thresholds, and local backgrounds differ, while weather, altitude, geomagnetic latitude, and surrounding buildings all influence the event rate. Even after corrections, a home network should not be presented as a replacement for specialized observatories or national radiation-monitoring networks.
Is CosmicWatch Safe to Build at Home?
CosmicWatch detects ionizing radiation but does not generate a hazardous radiation beam. Its count rate is not a calibrated radiation-dose reading, and the device cannot determine a person’s exposure without an appropriate conversion model and supporting measurements. The v3X contains low-voltage electronics and a plastic scintillator. Its scintillator can respond to ionizing radiation by producing light, but the scintillator itself is not radioactive.
The design is nearly open source, with build files and assembly documentation published in the project’s official GitHub repository. Even so, “buildable” is not the same as “plug-and-play.” Assembling the latest unit involves small surface-mount electronics, soldering, calibration, and testing. The older, simpler models are better suited to a first classroom build, while a teacher or lab technician should review any build plan and test equipment before students use it.
Because the project uses a noncommercial open-source license, anyone wishing to sell completed units or use the design commercially should review the license terms rather than treat the files as unrestricted public-domain material.
The Bigger Point: Real Science Does Not Always Require a Billion-Dollar Lab
CosmicWatch will not replace facilities such as CERN, IceCube, or the large dark-matter experiments used to search for extremely rare events. It does something more practical. It lowers the first barrier between curiosity and measurement.
A student can assemble the electronics, collect data, identify background events, compare altitude and pressure, run a coincidence experiment, and confront the uncertainty in a result. Those skills transfer to particle physics, astronomy, medical imaging, radiation monitoring, and many engineering fields.
CosmicWatch also illustrates a broader principle of scientific innovation. The breakthrough is not simply that the detector is cheap. It is that the instrument is modular, documented, and accurate enough to answer meaningful questions. Thousands of small, well-calibrated instruments could collectively reveal patterns that no one could study with a single detector alone.
It is a box of parts that can count the universe. For a new student, that may be the most exciting way to begin.
