This post describes two related projects: the astro-lab@home, published in CAPjournal in 2022 with Alexander Küpper and André Bresges [[1]]; and its successor, the astro-lab@school, published the same year in Astronomie+Raumfahrt [[2]]. Both grew from the same question: what does astronomy education look like when you cannot bring students into a lab?


What the astro-lab Was

Before the pandemic, the astro-lab at the University of Cologne was a student laboratory focused on extrasolar planets. School groups — mostly secondary school students — came in and worked through a set of analogy experiments: how do you detect a planet you cannot see? How do you infer its size, its orbit, whether it might be habitable?

The pedagogical bet was that exoplanet research, precisely because it is headline-generating and genuinely open-ended, could counteract the motivational slump in physics that tends to set in around middle school, the grades 7 to 10 for which the lab’s exoplanet series was written [[3]]. The context — life in the universe, habitable worlds, the possibility of something out there — did a lot of the work that no abstract force diagram could do.

The experiments themselves were analogy experiments: a lamp standing in for a star, a sphere on a track standing in for a planet. The key measurement was the transit: when the “planet” passed in front of the “star”, the light sensor registered a dip. Students measured the dip, estimated the ratio of areas, connected it to radius, and got a number that meant something. The number was not precise. It did not need to be. It was real.


Spring 2020

In March 2020, schools shut down, and the University of Cologne followed. Visits to the astro-lab were cancelled. The question the team faced — Alexander Küpper, André Bresges, and I — was not whether to do something but what was actually feasible.

In the JIMplus 2020 survey (a representative online-panel sample of 1,002 German-speaking school students aged 12 to 19, fielded 2–6 April 2020), 16% of the students reported using video conferences with teachers or their class, and 30% said they had received assignments at the start of the shutdown and had hardly had contact with their teachers since [[4]]. Those figures describe that survey and period, not every German learner. The infrastructure was not there, the habits were not there, and the expectation that students had the materials and equipment for a physics lab at home was not warranted.

Smartphones were widespread, but access was not universal and did not guarantee that a required sensor was present or exposed to an app.

Smartphone models expose different combinations of sensors, which may include ambient light, acceleration, rotation, pressure, and magnetic-field sensors. The app phyphox, developed at RWTH Aachen, makes those sensors accessible with a clean interface designed for use in education [[5]]. If the sensor hardware was already in students’ pockets, the lab setup problem became: what household materials can stand in for the rest of the apparatus?


astro-lab@home: Bringing Science to the Sofa

The astro-lab@home project adapted the original lab experiments for home use with smartphones and everyday materials [[1]]. The core transit experiment — measuring the dip in light caused by an opaque object passing in front of a lamp — was demonstrated with household equipment and compatible phones. The publication is an implementation report, not a multi-site replication study. A table lamp, a ball on a string, and a smartphone positioned beneath the lamp gave you the raw data. In the version photographed for the paper, a globe-shaped garden light is the star, a styrofoam ball on a toothpick is the planet, carried around the lamp by hand, and a smartphone held upright at the lamp’s side is the detector; each orbit gives one dip. phyphox recorded the light curve in real time and, used as an optical stopwatch, computed the transit depth, the transit duration and the length of the planet’s “year” [[1]].

The transit was the first of seven activities in the paper; the others cover albedo with a reflective “planet”, tidal locking, atmospheric spectra in an app, the habitable zone and the greenhouse effect with an external sensor, and the air pressure in home-built spaceship “prototypes”, for which the paper recommends an external sensor to protect the phone [[1]].

We designed the setup to be flexible enough to work with what students actually had. The default used an ambient-light sensor on compatible Android devices. The phyphox App Store build cannot access the iPhone’s physical light sensor, so the paper suggested an external sensor, a TI CC2650 SensorTag or similar [[1]]; the phyphox FAQ still says that the iPhone’s sensor “cannot be accessed by our app” [[6]]. Cost, compatibility, calibration, and cross-device consistency depend on the chosen hardware and must be checked rather than assumed.

The resulting package was not just an equipment list. We developed accompanying materials that explained the physics (why does a transit produce a specific shape of dip rather than a sharp cutoff?), connected the analogy experiment to the real science (how does this scale up to the actual transit photometry done by TESS and Kepler?), and offered scaffolding at different levels of independence. The experiments went onto astro-lab.app, each loadable into phyphox from a QR code and accompanied by further information and instructional videos [[1]].

The project was published in October 2022 in the IAU’s CAPjournal, a journal aimed at communicators and educators in astronomy, as a “Resources” article [[1]]. The audience was intentionally broad: teachers looking for accessible classroom activities, outreach organisations trying to reach students at home, curious individuals who wanted to do something real with their phone. “Bringing science to the sofa” was the headline, and that was genuine. The experiments worked in a living room; the paper’s photographs have the sofa in the background.


Why the Dip Has Sloping Sides

The phone does not see a planet. Its light sensor reports illuminance, in lux on Android [[6]]: the light arriving per unit area from every part of the lamp in its view. Seen from the sensor, a globe of radius $R$ whose centre is a distance $D$ away fills a disc of angular radius $\alpha$, with $\sin\alpha = R/D$; a ball of radius $r$ at distance $d$ fills one of angular radius $\beta$, with $\sin\beta = r/d$. While the ball’s disc lies wholly inside the lamp’s, it hides the fraction

$$ \delta \approx k^2, \qquad k = \frac{r/d}{R/D}, $$

of a uniformly bright lamp’s light (exactly, for a ball on the axis of a sensor that faces the lamp and has an ideal cosine response). Here $k$ is the ratio of the apparent radii (for small angles, $\beta/\alpha$), not of the physical ones, so the same ball dims the lamp more the closer it is held to the phone (Fig. 1).

Figure 1. The closer the ball is to the phone, the deeper the dip and the shorter its floor. Top: an illustrative geometry (the paper gives no dimensions) with lamp, balls and distances to scale: a globe of radius 15 cm centred 60 cm from the sensor ($\alpha = 14.5^\circ$) and a 2 cm ball at three distances; $k = 0.8$ is included to show the near-V shape. Bottom: overlap of two discs for a uniformly bright lamp, no room light, and a straight crossing through the lamp’s centre line at fixed distance and constant speed; offsets in apparent lamp radii; ingress and egress drawn thicker. An ideal cosine-response sensor changes the curves by less than 0.01. Computed, not measured.

The ball’s disc has a size, so the dip is not a step. It begins when the ball’s silhouette touches the lamp’s edge, deepens while the silhouette slides onto the disc (ingress), stays level while it lies wholly inside, and recovers as it leaves (egress). For a straight crossing through the lamp’s centre line at fixed distance and constant speed, the level part takes a fraction $(1-k)/(1+k)$ of the dip’s duration: two thirds at $k = 0.2$, a ninth at $k = 0.8$. A ball held close to the phone gives a deep, nearly V-shaped dip, and at $k \ge 1$ it can hide the whole lamp, at which point the depth stops measuring its size. A ball carried by hand around the lamp is nearest the phone at mid-transit, so $d$ changes during the crossing and the floor need not be flat; the pointed dips in the example curve printed with the paper’s transit photograph are consistent with that. Room light adds a constant to every reading and makes each dip shallower relative to the baseline.

A star is darker towards its limb, which rounds the bottom of a real transit. An ideally diffusing globe lit from its centre is not: the glass is evenly lit, and a perfectly diffusing (Lambertian) surface has the same radiance in every direction, so its disc is uniform. Real opal glass, the lamp’s base and the phone’s angular response can each depart from that by amounts not measured here.

The dips are large enough to see with a phone: the uniform-disc depths of 4% to 64% in Fig. 1 compare with 1.06% for Jupiter crossing the Sun and 84 ppm for Earth, computed the same way from the IAU nominal radii [[7]].


What Came Next: astro-lab@school

When schools reopened and in-person teaching became possible again, the question was not simply “back to normal” but what the COVID period had actually taught us about the format.

The astro-lab@school, published in Astronomie+Raumfahrt in 2022, addressed that question directly [[2]]. Some design features from the home version were retained. The team judged the smartphone setup portable and useful for student control of the apparatus. The publications do not provide a controlled cost comparison or establish that using a personal device improves learning. The two papers appeared in the opposite order to the projects: the school paper in the first 2022 issue of Astronomie+Raumfahrt, and the CAPjournal report in October, which already cites it [[1]].

The astro-lab@school retained the smartphone-centred approach and adapted it for a school context: class sizes, time constraints, the reality of mixed equipment across a room of thirty students. The experiments from the home version were modified for group work and parallel execution. The scaffolding materials were reworked for the paced structure of a school lesson rather than the self-directed format of home use.

The result was not a reversion to the pre-pandemic lab. It was a hybrid: in-person group work, but with tools and methods developed for distributed individual use. The pandemic had pushed the format toward a different, more distributed design; “more robust” remains an engineering judgment rather than a measured outcome.


A Note on What Made This Work

The core technical contribution — smartphones as measurement instruments for analogy experiments in astronomy education — is described in more detail in a later publication in The Physics Teacher, which covers the experimental setups, sensor comparison, and pedagogical progression in a form aimed at an international teaching audience. If you want the how-to, start there.

What I want to note here is something slightly different: the role of context.

The astro-lab used exoplanets as a motivational context. The cited project record does not quantify motivation or compare exoplanets with other physics contexts, so I retain that choice as a design judgment rather than an effect claim. The COVID period showed that the team could implement the context without the university lab infrastructure. Students working on transit photometry with a lamp and a smartphone in their kitchen were doing the same thing, conceptually, as students at a benchtop sensor station at the university. The physical setup was different. The question was the same.

My narrower lesson is that this context survived one redesign. That does not prove that context-driven education is independent of material configuration.


For the curriculum unit that places these experiments in the context of the NRW Sekundarstufe I (lower secondary) physics syllabus, see Fremde Welten. For the air pressure / Mars experiment that grew from the same lab, see Mission to Mars.

Project and literature record checked through 2026-10-25.


References

[1] Spicker, S. J., Küpper, A., & Bresges, A. (2022). astro-lab@home – Bringing science to the sofa. CAPjournal, No. 31 (October 2022), 37–41; deposited in 2025 as Communicating Astronomy with the Public Journal 16(1), 37–41. DOI: 10.5281/zenodo.14986186

[2] Spicker, S. J., & Küpper, A. (2022). astro-lab@school: Ein Schülerlabor 2.0 für den Physik- und Astronomieunterricht. Astronomie+Raumfahrt im Unterricht, 59(1), 16–20.

[3] Küpper, A., & Schulz, A. (2017). Schülerinnen und Schüler auf der Suche nach der Erde 2.0 im Schülerlabor der Universität zu Köln. Astronomie+Raumfahrt im Unterricht, 54(1), 40–45.

[4] Medienpädagogischer Forschungsverbund Südwest (mpfs). (2020). JIMplus 2020: Lernen und Freizeit in der Corona-Krise. https://mpfs.de/studie/jimplus-2020/

[5] Stampfer, C., Heinke, H., & Staacks, S. (2020). A lab in the pocket. Nature Reviews Materials, 5(3), 169–170. DOI: 10.1038/s41578-020-0184-2

[6] phyphox. Frequently Asked Questions. Opened 25 October 2026. https://phyphox.org/faq/

[7] Prša, A., Harmanec, P., Torres, G., et al. (2016). Nominal values for selected solar and planetary quantities: IAU 2015 Resolution B3. The Astronomical Journal, 152(2), 41. DOI: 10.3847/0004-6256/152/2/41


Changelog

  • 2026-07-11: Bounded pandemic survey figures to their sample; corrected smartphone and iPhone sensor claims; and separated the published implementation from unmeasured replication, cost, robustness, motivation, and learning-outcome claims.
  • 2026-10-03: Corrected the bibliographic details in the reference list: the CAPjournal article is on pp. 37–41, not 12–17 (CAPjournal 31, October 2022); astro-lab@school is by Spicker & Küpper, Astronomie+Raumfahrt im Unterricht 59(1), 16–20, not 59(6); the 2017 Küpper & Schulz article has the title “Schülerinnen und Schüler auf der Suche nach der Erde 2.0 im Schülerlabor der Universität zu Köln”, pp. 40–45; and the phyphox entry, whose DOI does not resolve, is replaced by the Stampfer, Heinke & Staacks (2020) Nature Reviews Materials article that the CAPjournal paper cites. The 16% video-conference and 30% “hardly any contact” figures come from the JIMplus 2020 survey (mpfs, n = 1,002, 2–6 April 2020), not from the CAPjournal article, and are now attributed to it.
  • 2026-10-25: Added the CAPjournal paper’s version of the transit setup (a globe-shaped garden light, a styrofoam ball on a toothpick carried around it, a phone held at the lamp’s side) alongside the table lamp, ball on a string and phone beneath the lamp; added from the paper the phyphox stopwatch, the suggestion of a TI CC2650 SensorTag for iPhones, the six other activities and the distribution through astro-lab.app, and quoted the phyphox FAQ on the iPhone sensor. Added the grades 7 to 10 of the lab’s series (Küpper & Schulz 2017) and cited Stampfer, Heinke & Staacks (2020) for phyphox. Added a section with Fig. 1: the depth is $k^2$ for the ratio $k$ of apparent, not physical, radii; ingress and egress take a share of the dip that grows with $k$; a hand-carried orbit changes $k$ during the crossing; a diffusing globe has no limb darkening; Jupiter and Earth give 1.06% and 84 ppm. Noted that astro-lab@school appeared in print first. Completed the CAPjournal reference (“No. 31”; DOI 10.5281/zenodo.14986186, deposited in 2025 as 16(1)); updated the JIMplus URL and sample description; numbered the references.