STORY: Olin Students Create Rare, Fully Functional Electromechanical Turing Machine
A team of five Olin College of Engineering undergraduate students have created their own electromechanical Turing Machine. And at a low-cost of under $250, it might be one of only three physical machines in existence.
One of the other machines is Mike Davey’s machine (currently in a museum), which served as the team's inspiration.
"We got on the web and researched until we stumbled upon a similar project implemented by Mike Davey more than 10 years ago, and we decided that this is what we wanted to re-create (but with our own customizations and low-cost)," says the team.
The Oliners took on this project for the Principles of Integrated Engineering (PIE) course in the fall of 2023. After being inspired by the theoretical concept of the machine - introduced by Alan Turing in 1936, the students decided to go ahead and build a real one themselves for their final project for the class.
"Why not just make a real one?"
While the end result is quite amazing, the journey to complete the Turing machine in approximately 6-weeks, was arduous.
Sparsh Gupta '26, one of the machine's creators and an Engineering:Computing major, shared that the overall process was "very challenging, but also rewarding," as it required the seamless integration of mechanical, electrical, and software sub-systems.
Evidence of the monumental challenge to build one is as of today, there are likely very few, if any, fully functional physical Turing machines in the world, as the concept of a Turing machine is primarily theoretical and used to study computation, not designed to be a practical computing device. Therefore, most machines are educational models demonstrating the concept, not actual working contraptions.
Not so, with these Oliners' machine.
Pictured: The team's final PIE project: A physical, electromechanical Turing machine.
The students had a simple objective: make a cool electromechanical Turing machine while learning new stuff and enhancing their skills as a productive team.
To bring their Turing machine into existence, the team started by breaking that final deliverable goal into bite-sized, iterative chunks, such as designing the write head, devising a model for image recognition, choosing actuators, or fabricating components. From there, they composed those sub-goals into groups by importance and scope, which they called sprint goals.
They ended up making three sprints of two weeks each. Within each sprint, they divided tasks amongst themselves, assigning deadlines and rating them based on priority. Using this system, they were able to remain more or less on track with their goals for the duration of the project.
- Go here to see the team's process in detail based on how they progressed through the sprints...
- Check out the team's journey and learn more about their machine on their website...
The Turing machine in action!
Watch the team's completed project in action via their YouTube video.
Embracing constraints leads to innovation
With less than $250, the team had to devise creative and effective ways to utilize the parts, sensors, etc. available to them. This constraint pushed them to consider working substitutes for sensors, parts and microcontrollers (such as using an ESP32CAM for symbol recognition). The team also narrowed their focus on implementing a machine capable of mostly simple computations (like accepting a string of certain format: palindromes, even 0’s even 1’s, etc. or operations like increment, addition).
During sprint 3, just days before the final demo, the team faced a major setback as several components in the machine failed, and they had to dismantle the entire prototype.
"At that point, we were convinced we’d never get back to where we were and feared the project was headed for failure. But we pulled through with countless late nights and an incredible team, as in the end, we didn’t just rebuild it—we created a version better than we ever imagined,” says Sparsh.