Octave: Making Sense of Sensor Data
The advancement of machine learning and the availability of Big Data have opened up the possibilities of data-driven analysis and decision-making across nearly every industry. One example is the Construction industry, which increasingly deploys various sensing technologies in the field, e.g., GPS, RFID, etc. Data from these sensors can provide insight into task planning, safety risks, and worker productivity. Construction specialists need to make sense of this data, but it is often difficult to do so without knowing a general-purpose programming language.
In this project, we are developing an environment called Octave (Observable Connections between Tables, Algorithms, and Visualization in an EUP) – an end-user programming environment for analyzing and visualizing spatiotemporal sensor data. Octave presents data both as a table and as a visualization and provides a small set of simple computational primitives. The views of the data are tightly connected to provide a responsive and intuitive environment. For example, clicking on a table row will highlight the corresponding datapoint in the visualization, and vice versa.
Octave will be deployed in a classroom setting and is intended to help Construction Engineering Management (CEM) students develop computational thinking skills while solving domain-specific problems.