System for and methods of calculating, predicting, and simulating safety scores for motor carriers
View Patent ↗The system provides real-time safety scores for motor carriers, accurately predicts future safety scores, and allows users to simulate various safety scenarios. Users can input data and perform complex calculations, ensuring accurate and timely updates of safety scores. The system's efficiency lies in accumulating and analyzing extensive data, leading to more precise predictions of future safety scores. The calculated, predicted, or simulated safety scores are presented on diverse devices such as desktop computers, mobile phones, or other compatible devices.
1 . A computer-implemented method executed by one or more processors of a network-connected server system, the method comprising: (a) retrieving, via an authenticated network session, a first dataset consisting of inspection and crash records for a selected motor carrier from an FMCSA Safety Measurement System (SMS) database corresponding to a published snapshot date; (b) retrieving, via an authenticated network session, a second dataset consisting of inspection and crash records for the same motor carrier from an FMCSA Portal, the second dataset including records dated after the snapshot date and not yet released to the SMS database; (c) normalizing the first and second datasets into a canonical violation record format and de-duplicating records based on report identifiers and state/date combinations; (d) for each violation record, automatically assigning severity and time weights by mapping a recorded violation code to an FMCSA codebook; (e) when a violation code in a record is inconsistent with a natural-language violation description in the same record, automatically correcting the violation code by selecting-using a string-similarity match to the codebook exceeding a threshold-a corrected code and assigning corresponding severity and time weights; (f) computing, for each safety category, a current safety measure in accordance with the FMCSA SMS methodology using the assigned severity/time weights and an exposure measure, and computing a percentile rank for the carrier relative to a peer group defined by power units and inspection count; (g) predicting safety measures and percentile ranks for a plurality of future months by estimating numbers of clean inspections, violations by severity, and crashes using the carrier's history combined with peer-group monthly averages, and recomputing the safety scores for each future month; (h) detecting whether a current or predicted percentile rank crosses a predefined threshold and, in response, generating an alert and causing display of a drill-down breakdown of the computations in an interactive user interface; and (i) storing the normalized datasets and computed values in a non-transitory computer-readable medium.
2 . The method of claim 1 , wherein a first source of the one or more electronic sources is the FMCSA SMS database containing public records for motor carriers.
3 . The method of claim 1 , wherein a second source of the one or more electronic sources is the FMCSA Portal containing private records for motor carriers.
4 . The method of claim 1 , wherein a third source of the one or more electronic sources is police and weight stations sending inspection reports to motor carriers, including inspections, violations, and crashes reported to a motor carrier but not yet recorded at the FMCSA Portal by police and weight stations.
5 . The method of claim 1 , wherein a fourth source of the one or more electronic sources is data maintained by motor carriers.
6 . The method of claim 1 , wherein a fifth source of the one or more electronic sources is data maintained by service providers to motor carriers.
7 . The method of claim 1 , wherein a first methodology of the one or more methodologies for calculating safety scores is the Safety Measurement System (SMS) Methodology as published and updated from time to time by FMCSA.
8 . The method of claim 1 , wherein a second methodology of the one or more methodologies for calculating safety scores is the Driver Safety Measurement System (DSMS) Methodology as published and updated from time to time by FMCSA.
9 . The method of claim 1 , wherein a third methodology of the one or more methodologies for calculating safety scores is the Inspection Selection System (ISS) for Compliance Safety Accountability (CSA) Algorithm as published and updated from time to time by FMCSA.
10 . The method of claim 1 , wherein the computing the safety scores further comprises an automatic assignment of severity weights to violations received from the FMCSA Portal based on violation codes recorded by police officers in inspection reports.
11 . The method of claim 1 , wherein the user interface further comprises graphical elements allowing users to review and edit automatically assigned severity weights to violations and to correct codes of violations misprinted by police officers that prevent automatic assignment of the weights.
12 . The method of claim 1 , wherein the user interface further comprises graphical elements allowing users to enter data from nonelectronic sources manually.
13 . The method of claim 1 , wherein the user interface further comprises graphical elements allowing users to add and remove months for predicting and simulating future safety scores.
14 . The method of claim 1 , wherein the user interface further comprises graphical elements allowing users to review and change parameters used to compute predicted safety scores, including quantity of power units, miles, and carrier segments for prediction and simulation months.
15 . The method of claim 1 , wherein the user interface further comprises graphical elements allowing a user to change severity weights of violations and crashes being appealed to FMCSA to simulate their effects on the safety scores.
16 . The method of claim 1 further comprises predicting numbers of probable inspections, violations, and crashes for months of a period to be predicted and simulated based on average monthly numbers of inspection reports with no violations, violations with different severity weights in safety categories, and crashes per carrier among carriers having the same number of power units and inspections within each safety category.
17 . The method of claim 1 , wherein the user interface further comprises graphical elements allowing a user to change quantities of predicted inspections, violations, and crashes, their distribution over time, and severity weights.
18 . The method of claim 1 , wherein the user interface further comprises graphical elements allowing a user to calculate and save multiple scenarios of the predicted and simulated safety scores for the same motor carrier independently from other users.
19 . A non-transitory computer-readable medium storing instructions that, when executed by one or more processors, cause performance of the method of claim 1 .
20 . The method of claim 1 , wherein the string-similarity match comprises computing an edit-distance between the violation description and codebook entries and selecting the lowest-distance entry above a confidence threshold, and presenting the proposed correction for operator confirmation in the user interface.
21 . The method of claim 1 , wherein the peer group is determined by both (i) quantity of power units and (ii) number of relevant inspections in the category, and the percentile computation excludes months lacking exposure.
22 . The method of claim 1 , further comprising applying hysteresis to the threshold detection to avoid repeated alerts due to minor fluctuations.
23 . The method of claim 1 , wherein predicting the future counts uses a Poisson or seasonal average model blended with peer-group monthly averages.
24 . The method of claim 1 , wherein the user interface permits saving multiple scenarios per carrier and later comparing each scenario to actual scores as they are recomputed from Portal data.