Development of a Diagnostic System for Air Brakes in Autonomous and Connected Trucks (​04-100)

Project Description: This data was collected for the project “Diagnostics Using Machine Learning for Air Brakes in Commercial Vehicles”. This data was collected in June 2023 in Dr. Swaroop Darbha’s lab Doherty 301A, Texas A&M University, College Station, Texas. The setup described in the final report was used to collect the data. Data Scope: This data describes the pressures in the rear and front chambers of the air brakes in a typical commercial vehicle under hard braking. The data also contains pedal travel(mm) and air leakage velocity (ft/min) from the system. This data was used to build a machine learning model to predict the amount of leakage in the system, given the pressure traces of any chamber under hard braking. All the data reported is time series data and there are 23 tests performed under various different conditions. The files are named according to each test scenario, as “Dataleak.csv”. The data given is also labelled in case of any discrepancies. Data Specification: There are 4 Rear Chambers and 2 Front Chambers in the experimental setup. The data given has been labelled accordingly as RearChamber1,RearChamber2 etc. The time data is given in seconds and the zero point is calibrated to the test start point. The pedal travel is given in mm. The Leak Velocity is given in ft/min. Finally, a table named “dataset.csv” is included which contains derived data from all 23 scenarios. This table was used to build the machine learning model.