Software Glitch & Lack of App Testing Cause Major Delay in Iowa Caucus Results
On February 3, 2020, the 2020 US Presidential Election season kicked off with the Iowa Caucus. Typically, results are recorded and released within the same day; but this year, results were extremely delayed due to the implementation of a new application. Chairman of the Iowa Democratic Party Troy Price assured that data was being recorded accurately, but a coding issue was causing the data to be only partially reported. Price also made clear that this was not a security issue and there would be no impact on the accuracy of the data. The reporting fiasco caused a backlash by numerous people in the Software Testing industry after news broke that the app had not been security tested or vetted for statewide use; Homeland Security’s Cybersecurity and Infrastructure Security Agency had offered to test the app, but the app’s developers, Shadow Inc., declined.
An Out-of-this-World Software Error
For more than 7 years, the Mars Curiosity Rover has been navigating the red planet. There have been a few issues throughout its journey, mainly because the Curiosity was only designed to last a couple of years; but, as of mid-January, it’s found itself frozen in place. Because Mars is 183,390,000 miles away, NASA cannot directly control the Curiosity in real-time, but rather through batches of commands that it receives and carries out; as it carries out tasks, it stores this attitude data in its memory—but something went wrong one day. The Curiosity lost track of its orientation and froze itself in place in order to prevent damage. It is still in contact with NASA here on Earth, though, so now we’ll have to wait and see if NASA can unfreeze our interplanetary robot.
A Major Flaw in Tesla’s Autonomous Driving Sensors
Tesla’s Autopilot system is one of the leaders in the autonomous car movement. For $8,000, you can upgrade your Tesla to Full Self-Driving Capability, which allows the car to take you from point A to point B with no driver intervention thanks to countless sensors and image-recognition systems. However, a Ph.D. student in Israel set out to challenge this system—and was successful. By using a projector to display false road signs and lane markings, he was able to trick the system into following these false projections. This is simply because AI image recognition just does not yet compare to that of humans. The outcomes of this experiment highlight the necessity to integrate defensive designs into semi-autonomous driving systems, as well as the fact that fully self-driving cars may be further away than we previously thought.