Use of Machine Learning to Predict the Injuries of the Occupant of a Vehicle Involved in an Accident Maruti Suzuki India, Limited

Format:

Book

Conference/Event

Author/Creator:

Howlader, Ashim, author.

Conference Name:

Symposium on International Automotive Technology (2021-09-29 : Pune, India)

Language:

English

Physical Description:

1 online resource cm

Place of Publication:

Warrendale, PA SAE International 2021

Summary:

As per the 2018 MoRTH accident report, there were 467,044accidents, out of which 137,726 were fatal which resulted in151,417 fatalities. In order to get an idea of the reasons forinjuries and estimate the benefits of any intervention, amathematical model should go a long way. This study is aimed at thedevelopment of such a model to predict the injuries sustained bythe occupants of an M1 vehicle. We used a detailed accidentdatabase of 'Road Accident Sampling System India' (RASSI).RASSI, since 2011, has been collecting traffic accident datascientific across various locations in India. In the data, theoccupant injuries are classified as No injury, Minor, Serious andFatal We used the data of about 4700+ M1 occupants for the study& used almost 40 input parameters to determine the outcome.Based on the data, an algorithm was developed with an overallaccuracy of about 67%. The parameters represented human,infrastructure, and environment. In 67% of the cases, the injurieswere accurately predicted. In 14 % of the cases the predictedinjuries were one level above than actual id est for example in casethe actual injury was minor the model predicted it as serious weterm this as +1 shift error. Likewise, 11% of the time the modelpredicted injury one level lower than the actual id est for exampleif the actual injury was of a serious nature, the model predictedit as minor. These can be termed as -1 shift errors. But if wecombine ±1 shift errors and the 0 errors the accuracy increases to92%. The model can be used as a first step towards accessing theeffectiveness of an intervention. Post this more expensive fieldtrials may be carried out

Notes:

Vendor supplied data

Publisher Number:

2021-26-0003

Access Restriction:

Restricted for use by site license