The Therac-25 Case


The objective of this discussion is to introduce the Therac-25 case, explain what happened in this historical event, discuss the ethical responsibilities of such individuals and organizations and at the end provide my viewpoint about this historical event.


The way as the software becomes more complex and usable, this is very common that scientists introduce the use of software to the medicine. Between June 1985 and January 1987 disasters occurred with a system named the Therac-25, which was a computerized radiation therapy machine. In the related period, six known accidents occurred with massive overdoses applied to the patients (Leveson & Turner, 1993).

As stated by the Leveson & Turner, 1993, the Therac-25 software was developed by only one person, using the assembly language, and it took many years to be completed. However, there wasn’t much documentation.

The technical aspect of the Therac-25 failure is that that software controlled by the operator if the operator used to trigger the buttons of the controller board slowly, the machine would operate fine, because while the development of the Therac-25 machine it was tested with slow command inputs. However the operators of this machines after operating thousands and thousands of patients became familiar with the controller board, they got used to the machine and used to type very and very faster, what used to trigger the bug, and the bug started to send massive radiations over the patients. (Udacity – Software testing Course, 2017).

In my viewpoint, it is more of a software testing failure, but of course, as software testing is part of software engineering, it needs to be considered an engineering failure.


In spite of having to know that software developed in the seventies and to do not have been properly tested, is a big failure in the procedures of the development of the critical health machine. But a machine of such complexity in my viewpoint, to be developed by a single person without documentation is also a big mistake by the company who created it. Unfortunately, the software engineering methods and studies we have today weren’t available at that time.

Just by the factor of the software to have minimal or none documentation puts all the responsibility to the developer: How can we build something we don’t know how it works since assembly code is very close to the machine code and all the other equipment were handled by this single software? How could understand the error codes provided by the software since there is no documentation?

If the error code showed that the problem was related to the operating timing while using the control board, the software could be easily repaired and the machine could continue to be used and to save lives.


Since technology is under development and it is not certainly free of problems and mistakes. According to the mistake evidenced in this documentation, in my opinion, it is not a case of cyberethics nor computer ethics, but a concern of professional ethics. I purely believe that because the operators of the Therac-25 had to be in a constant training environment of the use of such machines. Since the first case, that should be evidenced what was the mistake happened, what was the cause of that and why should we continue operating such machine since the first people who died because of the mistake which happened.


Spite of an engineering mistake mixed to an operative mistake, the development of software engineering and the advances of the medicine has, of course, much to improve the social quality of life in terms of the health. I consider many mistakes may have caused the Therac-25 case, but of course evicted. Spite of all the factors, it can be considered a software engineering failure, since the program was waiting for the perfect use of the software, the software had no documentation and the problems could not be fixed since it had no feedback of what was going on and could not help in the fix of such problem.


N.G. Leveson, C. T., 1993. An investigation of the Therac-25 accidents. IEEE Computer Society, 26(7), pp.18-41.

Udacity, 2017. YouTube. [Video Online] Available at: [Accessed 27 August 2017].

Published by Ademir Constantino

Software Engineer

%d bloggers like this: