Can Automobile Insurance Telematics Predict the Risk of Near-Miss Events?

Abstract

Telematics data from usage-based motor insurance provide valuable information -including vehicle usage, attitude towards speeding, time and proportion of urban/non-urban driving -that can be used for ratemaking. Additional information on acceleration, braking and cornering can likewise be usefully employed to identify near-miss events, a concept taken from aviation that denotes a situation that may have resulted in an accident. We analyze near-miss events from a sample of drivers in order to identify the risk factors associated with a higher risk of near-miss occurrence. Our empirical application with a pilot sample of real usage-based insurance data reveals that certain factors are associated with a higher expected number of near-miss events, but that the association differs depending on the type of near-miss. We conclude that nighttime driving is associated with a lower risk of cornering events, urban driving increases the risk of braking events and speeding is associated with acceleration events. These results are relevant for the insurance industry in order to implement dynamic risk monitoring through telematics, as well as preventive actions.

Near‐miss telematics in motor insurance

“…Telematics offers the possibility of recording this type of near‐miss data in everyday transportation. Near‐misses in the context of automobile insurance have recently been investigated by Guillen et al (2020) in a sample of drivers with an OBD designed to collect information on driving indicators through continuous measurement. They found that risky driving behaviors such as driving fast or at night are associated with the occurrence of near‐misses, but they did not apply this result to actual claims and they did not discuss insurance ratemaking.…”

Driving Risk Assessment Using Near-Miss Events Based on Panel Poisson Regression and Panel Negative Binomial Regression

“…Conventional generalized linear models discern the correlation between influencing factors and claims or accidents in frequency and severity models [ 9 , 24 , 25 , 35 ]. However, the study of near-miss events even when there is a lack of information on claims and accidents should not be ignored [ 2 , 15 ]; on the contrary, since near-misses are more frequent than accidents and are positively associated with them, they can be considered a good alternative for risk modeling for driving risk assessment [ 1 ]. Compared with previous studies, this study not only conducts regression on the summary data set to model and analyze the factors causing near-miss events, but also conducts panel data regression on the panel data set to consider individual effects and time effects.…”

“…Lastly, both harsh acceleration and harsh deceleration are near-miss events that compromise driving safety and fuel economy. Based on previous research experience [ 1 , 2 , 37 ] and the filter analysis of the extreme values of this data set by box graph method, 6 m/s is determined as the filtering threshold value of harsh acceleration and harsh deceleration. Figure 1 shows that near-miss events are all non-negative integers.…”

“…Fortunately, the advent of the Internet of Vehicles (IoV) offers a better solution to this problem, using near-miss events to identify driving risk. Near-miss events ultimately provide information that can lead to actuarial premium calculations in the auto insurance industry [ 1 , 2 ].…”

From pool to profile: Social consequences of algorithmic prediction in insurance

“…Assuming, for example, that risk exposure is directly proportional to the number of miles travelled by car (as is the case with PAYD policies) is questionable because as the number of miles driven increases, so does the driver’s experience (the so-called ‘learning effect’), with the result that the number of accidents decreases while the number of good drivers increases in relation to the frequencies expected if the ratios were directly proportional (Guillen et al., 2019a). At the same time, some actuarial mathematicians suggest to consider also the correlation of telematics data with so-called ‘near-miss’ events, i.e., narrowly avoided accidents, as the latter would be strictly correlated with the risk of being involved in future car accidents (Guillen et al., 2019b). The underlying reasoning would be that those who often come very close to having an accident should be treated as those who already claimed, although there has not (yet) been any accident.…”