Social Issues and Ethics in Computer Science and Engineering

Introduction

Therac –25 is a medical linear accelerator that was developed by AELC .A linear accelerator (linac) is a particle accelerator, a gadget that increases the energy of electrically charged atomic particles. Linacs are use mainly in hospitals to treat cancer patients .During treatment a patient is exposed to beam of radiation in doses designed to kill a malignancy.(Grolier, 1985)

The Software Engineering Code of Ethics and Professional Practice is a practical frame- work for moral decision-making related to problems that software engineers may encounter. (Quinn, 2013)

Between June 1995 and January 1987, six patients were seriously injured and some killed by poor administration of radiation from the Therac-25 medical linear accelerator. This paper therefore seeks to explore the causes behind the accidents, the software bugs that were associated with the machine. In addition the paper will also cover some of list the clauses that are violated in the code of ethics of software engineering and explain how they relate to the action or inaction that led to the overexposure incident

Technical errors in Therac-25 software One of the major weaknesses that is associated with Therac-25 software was in the lack of formal testing procedures. As results certain errors remained in the software as the product got distributed to the consumers. One of the errors that were common with the software was radiation overexposure which can be attributed to interface errors on part of the machine. Another error involved the entering of treatment data by the machine operator. In addition once the machine’s magnet are set, there is no test performed to double check the treatment information entered matches how the magnets are set (Jacky, 1989) This was in total violation with the Software Engineering Code of Ethics and Professional Practice which calls for software engineers to approve software only if they have a well-founded belief that it is safe, passes appropriate tests, and does not diminish quality of life. Another major discrepancy on the software that played a big role in the death of the six patients was the nature of the real time system that had been installed in the system. When the machine is in use a variable called class 3 is usually set to a non-zero value, whenever the operator is testing the position of the light beam, this variable usually changes. Once testing is done the variable is usually set to zero while the radiation beam is allowed to pass. Despite this class 3 was stored in one byte memory. The effect of this was that every 265th increment would results in the value of zero assigned to it. Looking at the case study of the therac-25 incident, the operator pushed the set button at the exact moment that class 3 rolled over to zero. Consequently a full prescription beam was released without any of the beam flatteners ( 1993). This is another malfunction that would have been avoided if the engineers would have conducted a bug test to the machine before they had released the machine to the users, AELC was responsible for many of the factors that led to the Therac-25 incident and poor engineering methods were one of the major failures. For instance AELC violated another code of ethics that govern engineers when they did not responsibility for, correcting, and reporting errors in the software even after the software had stalled. Moreover only one programmer had been assigned to creating the machine’s complex real time software. Software specifications and testing criteria had not been laid down and very little software and system testing was performed (Jacky, 1989). Looking at the Therac-25 cases study it is apparent that poor engineering practices played a big role in the malfunctioning of the software, As a result if, I were part of the development of a product with an embedded system such as Therac-25, in order to prevent the occurrence of such an incident I would ensure that safety of machines is manufactured into the product, not after we receive complaints from the users. Moreover I would ensure that I have assigned a team of engineers when it comes to building a real time feature for such a system. This way minimal or no errors can be made during the design process that might affect the machine later during usage by end users,

References

Quinn, M.J(2013)Ethics for information age. Addison Wesly.PDF

Gowen, Lon D. and James S. Collofello.(1994) Software Safety and Preliminary Hazard Analysis."Professional Safety .

Jacky, J(1989). Programmed for Disaster. The Sciences .