At a dinner during IROS2026 in Pittsburgh I was sitting next to Paul Oh and we got chatting about the famous “ball and beam apparatus” which is a staple of undergraduate control labs. In particular where it had come from. I’d used one during my masters degree and here are some pictures of it from my thesis — yes, actual photographs from a film camera printed onto an A4 page in a darkroom!
In this implementation the ball rolls along two nichrome wires and act as the wiper of a potentiometer. The wires were a real problem, the contact was not perfect so the position signal was very noisy, particularly if the control input excited a vibration in them. The wires are long and if they were a bit slack could pinch the ball and stop it rolling easily. Finally, the beam could flex or twist laterally which could cause the ball to fall off.
The apparatus first gets mentioned in a paper by Peter Wellstead in 1978:
Wellstead PE, Chrimes V, Fletcher PR, Moody R, Robins AJ. The Ball and Beam Control Experiment. International Journal of Electrical Engineering & Education. 1978;15(1):21-39. doi:10.1177/002072097801500107
The lab I worked at (my first job 1981-83!) bought one of these, and I’m pretty sure it arrived in 1982. The lab was led by Prof John Anderson who came to Australia from the UK via Waterloo. The UMIST Control Systems Centre (under Howard Rosenbrock, Alistair MacFarlane, and Wellstead) in Manchester was the hub of controls research in the 60s and 70s — my prof (state space techniques for modelling steam boilers, power plants and grids) worked there in the late 60s.
Gemini says that UMIST licensed it to TecQuipment and that “It was originally released around 1980–1981 as the TecQuipment CE6 Ball and Beam Apparatus (later updated and known today in their catalog as the CE106)”. And that’s exactly what’s on the nameplate in the picture. I’m guessing we had a very low serial number unit. Wellstead actually wrote the manual shipped with the machine:
Wellstead, P.E. (1981), CE6 Ball and Beam Apparatus Manual, TecQuipmnent Ltd., Long Eaton.
But there are a couple of interesting side stories here. Firstly, the source of the idea and this is revealed at the end of the Wellstead paper:
ACKNOWLEDGEMENTS
The ball and beam experiment did not originate with us. The idea was brought to our attention in the laboratories of the Division of Automatic Control, Lund Institute of Technology, Lund, Sweden.
We acknowledge the advice and help given us by Professor K. J. Astrom and his colleagues of Lund Institute of Technology. In particular the comments made by Johan Wieslander were most helpful. During the design, construction and testing of the equipment many helpful suggestions were made by Mr. P. Bowler and R. Shelley of the Electrical Engineering Department, U.M.I.S.T. Postgraduate students from the Control Systems Centre who contributed significantly are J. M. Edmunds and B. H. Thiang.
Johan Wieslander was one of Karl Åström’s early doctoral researchers and senior staff engineers at Lund (completing his thesis in 1971 on real-time adaptive control). Around 1970–1972, Wieslander and the laboratory engineers at Lund built several custom, prototype bench rigs connected directly to the department’s PDP computer interfaces. Among these was the prototype known in the lab as the ball-on-track or kula på ränna:
- The Original Lund Prototype: Unlike Wellstead’s later dual-wire approach, early prototypes at Lund experimented with slotted tracks, simple rolling cylinders, and different position-sensing schemes (including early resistive strips and optical phototransistor arrays).
- The Role in Real-Time Software: The Lund lab used this rig specifically to develop and showcase interactive computer-aided control systems and real-time executive software. Wieslander was a key developer of early interactive software packages (like INTRAC) used to log data, fit transfer functions, and switch digital feedback loops on the fly.
- Rather than publishing the hardware design as an educational apparatus in its own right, Lund treated it primarily as an in-house laboratory demonstration and software test rig, describing it in departmental technical reports (LUTFD2/TFRT series) and student lab manuals rather than broad journal papers.
During the 1970s, UMIST and the Lund Department of Automatic Control (under Åström) had an exceptionally close transatlantic/European exchange:
- Researchers, postdocs, and professors constantly shuttled between Manchester and Lund for sabbaticals, symposia, and IFAC working groups on self-tuning and identification.
- Peter Wellstead visited Lund in the mid-1970s, saw Wieslander’s ball-on-track experiment running off the Lund real-time systems, and recognized that it was the ideal physical embodiment of cascaded position/velocity dynamics.
- Wellstead brought the concept back to Manchester, where he and UMIST colleagues—notably P. Bowler and R. Shelley (who headed UMIST’s electrical power and machine workshops)—engineered a ruggedized, repeatable version. They introduced the conductive dual-wire track, where the ball itself served as a rolling potentiometer wiper, solving the sensor noise and mechanical friction issues that plagued early laboratory prototypes.
- Wellstead also brought in graduate students like J. M. Edmunds (who later became well-known for multivariable frequency-response CAD software) to formalize the state-space matrices and frequency-domain models for his 1978 paper.
Secondly, TecQuipment is a pretty interesting story in its own right — a university startup from 1958! It was founded by an interesting partnership between academia and precision manufacturing:
- Sir Joseph Pope – At the time, he was a Professor of Mechanical Engineering at the University of Nottingham (and later became Vice-Chancellor of Aston University). Pope was an influential engineering academic who recognized that university engineering education suffered from a lack of purpose-built, standardised laboratory teaching hardware.
- William Cope – A skilled, renowned precision engineer and clockmaker.
In the late 1950s, engineering departments that wanted practical experiments had to task their own workshop technicians with building one-off apparatus from scratch, which was expensive, time-consuming, and inconsistent. Pope teamed up with Cope to take university-level theoretical concepts (in structures, materials, fluid dynamics, and thermodynamics) and turn them into commercially manufactured, repeatable benchtop experiments. Cope provided the fine mechanical and instrument-making expertise, while Pope ensured the apparatus directly mapped to engineering curricula.
By the 1970s and 1980s, when control engineering exploded as a distinct discipline, they partnered with academic groups (notably UMIST and other UK control centres) to produce apparatus like the CE-series control rigs—including the ball and beam—which made their way into universities across the Commonwealth and worldwide.
Today there are many vendors of what we should probably call kula på ränna, as well as open source resources should you be minded to build one yourself. There is also the 2-dimensional version, the ball and plate apparatus.
Written by Peter and Gemini.