Research
QR Code Secret Data Scheme & Scanner
By: Dan LaChance & Greg Johnson
QR codes are an effective way of sharing information quickly and reliably, primarily because of their error correction capabilities. This error correction feature can be harnessed for secret data sharing by changing bits that the scanners will be able to correct. This secret data can compromise voting security by violating privacy by hiding voter identification in the ballot QR code, as outlined in VVSG 2.0 guidelines. Our team developed two proof of concept schemes, tested their effectiveness, and tested codes to analyze their number of errors. We concluded that any QR code with any errors should be treated with suspicion because an attacker may have tampered with it. In order to detect errors, it is important that every voting center use a tool to scan the number of errors in a code before extracting any data. Although many voting machines’ QR codes can withstand 25% error capacity, if human workers can not see what is causing a high amount of errors with their eyes, the code has most likely been tampered with. Common ink smudges and bad lighting will be recognizable to anybody but secret messages will not. It is important for voting centers to do so in order to better secure our elections.
Partial-Temporal Constraint in Constraint-Based Programming
By: Dan LaChance & Greg Johnson
Constraint Based Programming is a means of solving complex constraint satisfaction problems characterized by a tuple of variables, domains, and constraints that must be met in order to find a feasible solution. Our focus is on scheduling problems using constraint programming. One method for creating these constraints is temporal, a method that creates a schedule with strict ordering of all constraints; they must happen one after the other. In this study, we analyzed constraint based programming and how we can properly test and model less restrictive temporal constraints. Empirical study was done by producing a schedule for a FIRST robotics competition. In this competition, there are two repeated tasks: a five-minute competition that must be completed three times and a set of ten-minute practices, where all practices must be completed before the last competition. We used Google-OR Tools to effectively create two models that could be compared to each other. One model uses old temporal methods to create a solution, and the other uses a proposed relaxed temporal method to find a solution. Overall, the new, relaxed constraint was able to produce a shorter competition day because it was able to output solutions that made better use of the resources. Read More
