On-board Diagnostics Research Articles

Framework for Building Low-Cost OBD-II Data-Logging Systems for Battery Electric Vehicles

With the electrification of transport (BEVs) and the growing benefits of smart vehicles, there is a need for a simple solution to perform real-time monitoring of the BEV and its battery for diagnostics and coordinated charging. The On-Board Diagnostics (OBD) system, originally designed for internal combustion engine cars (ICE), can be used to extract the necessary BEV data. This paper presents a developed framework for a low-cost solution to online monitoring of BEVs. A Raspberry Pi Zero W, along with other auxiliary components, was installed in two Hyundai Ioniq Battery Electric cars to communicate with the vehicles via the OBD-II port. A python script was developed to periodically request the vehicle data by sending various Parameter IDs to the vehicles and storing the raw response data. A web server was created to process the hexadecimal encoded data and visualize the data on a dashboard. The key parameters, such as the battery state of health (SOH), state of charge (SOC), battery temperature, cell voltages and cumulative energy consumption, were successfully captured and recorded, which can now facilitate trending for battery diagnostics and future integration with smart chargers for coordinated charging.

A fuel-consumption based window method for PEMS NOx emission calculation of heavy-duty diesel vehicles: Method description and case demonstration

Due to the lack of engine reference torque and the accumulated work of reference transient cycle, the work based window (WBW) method for portable emission measurement system test data processing cannot be used for vehicle emission assessment in the current on-board diagnostics (OBD) system in China. In this work, a fuel-consumption based window (FBW) method was proposed to imitate a WBW method procedure by using fuel consumption rate as an alternative parameter to scale the window so the entire procedure can be based on the attainable data items in the OBD system. Some key issues regarding converting WBW method to FBW method, including window separation, window average power ratio calculation and specific NOx emission conversion from mg/kg. fuel to mg/kW.h, were solved by linking the 100-km fuel consumption and the average vehicle specific power of China World Transient Vehicle Cycle test. The comparison between the FBW and WBW methods on the NOx emission calculation results shows that the number of all windows, the number of valid windows, and the thresholds for >50% valid windows are quite similar for WBW and FBW methods. The estimation accuracy of average power ratio for the FBW method depends on the value of transmission efficiency of vehicle driveline. The deviations of 90% specific NOx emission in mg/kW.h between the two methods are smaller than 6% for the cases investigated in the present work.

Experimental study on the performance and diagnosis of hydrothermally aged selective catalytic reduction over a copper-based catalyst

The conversion of nitrogen oxides (NOx) by hydrothermally aged copper (Cu)–based selective catalytic reduction (SCR) is an irreversible phenomenon. Tail exhaust NOx emission control and on-board diagnostics (OBD) monitoring are difficult with hydrothermally aged Cu–based SCR because the on-board NOx sensor has cross sensitivity to NH3 in the exhaust. In this study, the performance and diagnosis of hydrothermally aged SCR under real engine exhaust were studied. The influence of hydrothermally aged SCR on NH3 storage, tail exhaust NOx, urea injection, and OBD diagnosis were quantitatively analyzed under the World Harmonized Transient Cycle (WHTC). The results show that hydrothermally aged SCR had a great influence on the NH3 storage capacity. In terms of SCR conversion efficiency and OBD detection, compared with fresh SCR, hydrothermally aged SCR was more sensitive to space velocity than to temperature. Under different OBD monitored release conditions, the NOx conversion efficiency was greatly different; hence, the occurrence of a false diagnosis can be avoided through the selection of appropriate release conditions. These research results provide a reference for SCR size design, NOx control, and diagnostic strategies.

Определение затрат энергии на гидростатический привод мотор-вентиляторов охлаждающего устройства дизеля тепловоза ТЭП70БС по данным микропроцессорной системы управления

Purpose: Development, testing methodology for the definition of energy expenditure on hydrostatic drive of cooling device motor fans of diesel locomotive TEP70BS. Energy expenditures on drive of cooling devices motor-fans constitute significant part in the total volume of energy expenditures for auxiliary needs. When determining diesel cooling device power, picked for motor-fan drive, it is necessary to take into account the influence of motor-fan drive technical condition and diesel cooling system general technical condition. The given task can be solved by the way of statistical analysis of the information, being registered by TEP70BS locomotive on-board diagnostics subsystem when performing corresponding analytical calculations. At the same, it is necessary to account for the peculiarities of functioning of motor-fan hydrostatic drive under various loading modes. Methods: Statistical processing and analysis of measurement results. Analytical calculation of a hydrostatic system on the basis of processing of measurement results. Reduction of hydrostatic drive parameters to non-nominal modes. Results: Methodology of energy expenditure analytical calculation on the basis of statistical analysis of data on motor fan operating modes is proposed. Probable values of power, spent on motor-fan drive for various controller positions, are determined. It is shown that for controller nominal position, average power consumption for being considered sample of five locomotives will constitute 23.2 kW per two fan motors. Practical importance: The use of the presented methodology allows to raise significantly the determination accuracy for power, being picked for auxiliary needs and, as a result, to increase calculation accuracy for traction characteristics and fuel consumption characteristics of TEP70BS diesel locomotive.

Effect of Gear Usage on the Fuel Consumption of Light Duty Vehicle by Using On-Board Diagnostics (OBD) DataEffect of Gear Usage on the Fuel Consumption of Light Duty Vehicle by Using On-Board Diagnostics (OBD) Data

Transmission gear ratio usually affects transportation fuel usage during travelled on road. The manner of gear usage is one of the main factors affecting the fuel consumption of vehicles. By using the neutralizing driving data collection, the vehicle Toyota Hilux 2.5 liters has been used to conduct the experiment to determine the fuel consumption, based on the average vehicle speed, engine speed and gear usage. Some parameters were investigated, with a special emphasis on the relation between the engine speeds and vehicle speeds, also presenting an example of using the information provided by the OBD to determine the gear ratios. In the end, the gear steps and the way they are used while driving the vehicle were identified. It is found that each gear usage can have a specific impact on fuel consumption and the usage of first gear indicates higher average fuel consumption with 60.5% with 35.8 liters/100km compared with second gear. In conclusion, the fuel consumption is significantly determined by a proper gear selection

Protocols for Secure Remote Access to Vehicle Onboard Diagnostic Systems in Smart Cities

The onboard diagnostic system (OBD) is almost standard on every modern vehicle. By exploiting the availability of wireless OBD dongles, vehicles can be assessed by the transportation authority and drivers to ensure compliance with emissions regulations. A vehicle’s OBD system can be remotely accessed by a fog node to retrieve a variety of the vehicle’s internal status and sensor readings. Such information can be shared with the transportation authority for registration renewal without the need for visiting emission stations. However, such automation comes at the risk of malicious intrusion into the OBD system by an attacker who aims to gain access to a vehicle’s internal network and control the vehicle. This article tackles this serious issue and proposes protocols for mutual authentication and securing the wireless connection between the OBD dongle, fog node, and authority. The robustness of the proposed protocols is verified using the Automated Validation of Internet Security Protocols and Applications (AVISPA) toolset.

AI-based classification of CAN measurements for network and ECU identification

Due to the constantly increasing number of functions offered by a modern vehicle, the complexity of vehicle development is also increasing as a result. A first indication of this connection is provided by the number of ECUs (electronic control units) used in current development vehicles. Furthermore, each ECU also performs more functions and is not only electrically networked with the other ECUs, but also logically and functionally. On this basis, new cooperative functions are being developed, which are used for example for autonomous driving. In vehicle development, more and more test sequences (diagnostic scripts) are established for function testing of individual components, systems and cross-functional methods. Due to decentralization and the modular approach, modern development vehicles consist of different numbers of ECUs. The high number of ECUs in purpose and number poses a challenge for test creation and updating. The ECU software is also developed in cycles within the vehicle cycle. This results in a very high software variance. This variance leads to the fact that in the vehicle development with global test conditions works. Global test conditions at this point mean that more ECUs are included in the measurement procedure than are installed in the vehicle. The vehicle structure (control unit and its software version) is not known to the person performing the measurement. He relies on the fact that his ECUs are inside in the global measurement task. This means that the vehicle network architecture is uncertain, which can lead to errors during test execution. Since the ECUs that are actually installed in the vehicle are first determined during test execution, this results in a longer script runtime than would be necessary. To support the development engineer and prevent avoidable errors, the diagnostic system should configure itself as far as possible. This means that individually customized measurements for each vehicle should be calculated in the cloud and not the global measurement tasks. For a diagnostic system to be able to configure itself independently, the vehicle network structure must be determined in a first step. This can be done by a simple CAN measurement (measurementXY.asc). An AI is able to analyze this measurement and classify the occurring ECUs as well as CAN networks. For larger measuring devices with more than one CAN interface, the user who analyzes the measurement is interested in which CAN was connected. Here, the AI is suitable to determine the name of the network and the communicating ECUs based on the communication that runs over the bus. For this purpose, the AI classifies the number of communicating ECUs based on the time intervals at which messages are sent. In addition, the AI can be supported by a special diagnostic script (global.pattern) to determine the vehicle structure at the OBD (on-board diagnostics) interface with maximum accuracy. Three AI approaches are presented, all connected in series and passing results to each other (pipeline mode). First comes the AI that separates vehicle communication from diagnostic communication. Based on the vehicle communication, the network name can be determined. Based on the diagnostic messages, the ECUs can be determined.

Experimental Investigation on OBD Signal and PN Emission Characteristics by Damaged-DPF Types of 2.0 L Diesel Vehicle

A diesel particulate filter (DPF) is an exhaust after-treatment device designed to capture and store exhaust particulate matter, such as soot and ash, to reduce emissions from diesel-powered vehicles. A DPF has a finite capacity and typically uses a substrate made of ceramic material that is formed into a honeycomb structure. Diesel particulate filters play an important role in diesel-fueled vehicles. Failure to maintain these filters can have significant consequences for vehicles. In this study, we investigated the failure type in cordierite DPF substrates. In addition, we experimentally characterized the particle number (PN) emission and on-board diagnostics (OBD) signal of a 2.0 L diesel-fueled vehicle generated by three types of DPF failure (crack, melting, and hollow). Specifically, X-ray photography analysis of the cordierite DPF was performed. The PN and OBD signals were assessed via the KD-147 vehicle driving mode and measured using a DMS-500 (PN measurement device) and global diagnosis tool (GDS) scanner (OBD diagnostic device), respectively. X-ray photography was used to characterize the internal structure of the three DPF-failure samples. A key result was that the maximum value of the OBD data, including airflow mass, boost pressure, and VGT actuator, was distinctly different for each DPF sample. The exhaust temperature gradient for the normal DPF and crack-damaged DPF followed the KD-147 driving pattern. This was because there was no volume damage inside the cordierite DPF substrates. However, in the case of the hollow and melting-damaged DPF, the volume inside the cordierite DPF substrates was reduced or the time for the exhaust gas to stay in the DPF substrates was decreased. The melting-damaged DPF continuously emitted the largest number of nanoparticles (of the order of 109 #/cc). This was regardless of the vehicle driving speed in the KD-147 driving mode. Eventually, an OBD-based algorithm to determine whether a DPF is damaged was derived in this study.

Characterization of Gear Usage to Fuel Consumption of a Light Duty Vehicle from Direct Measurements by Using On-Board Diagnostics (OBD) Data

Transportation contributes a major share of the fuel usage from total vehicles in developing countries. The manner of gear usage is one of the main factors affecting the fuel consumption of motor vehicles. The presented results came from the tests conducted on Toyota Hilux 2.5 liters vehicle to determine the fuel consumption, based on the average vehicle speed, engine speed and gear usage collected from using the neutralizing driving data collection. Some parameters were investigated, with a special emphasis on the relation between the engine speeds and vehicle speeds, also presenting an example of using the information provided by the OBD to determine the gear ratios. In the end, the gear steps and the way they are used while driving the vehicle were identified. It is found that each gear usage can have a specific impact on fuel consumption and the lowest fuel consumption recorded under fifth gear with 9.75 liters/100km which reduction of 4.3% reduction compared with 4th gear. In conclusion, the fuel consumption is significantly determined by a proper gear selection.

Assessing heavy-duty vehicles (HDVs) on-road NOx emission in China from on-board diagnostics (OBD) remote report data

Mobile source emissions are some of the major causes of air pollution across the world and are associated with numerous adverse health impacts. China has implemented increasingly stringent emission standards over the past few decades, the most recent one being the sixth emission standard (China VI) first rolled out in 2019. The China VI standard places special emphasis on tightening limits for NOx emission and introduces remote monitoring of vehicles' On-board diagnostics (OBD) data to reduce emissions from diesel and gas-powered heavy-duty vehicles (HDV).This paper aims to establish a methodology to calculate HDV NOx emissions on an extended timescale, and under real-world operations. OBD data was collected and examined from 53 China VI diesel HDVs and 14 China V diesel HDVs, which found that the average NOx emission factors are 1.420 g/km and 3.894 g/km for the two samples respectively; this indicates that the implementation of China VI standard helps to significantly reduce NOx emission from HDVs. Combining the emission factors and calculated travel distances collected by OBD and vehicle sales data, the China VI emission standard is estimated to have resulted in 43,969 tons of NOx emission reduction by the end of 2020. With China announcing country-wide enforcement of the new standard in 2021, >1.7 million tons of NOx emission could be avoided by 2023 annually.

Trajectory Data Acquisition via Private Car Positioning Based on Tightly-coupled GPS/OBD Integration in Urban Environments

The explosive growth of road vehicles especially the private cars has brought unprecedented pressure to a series of problems in urban transportation systems, such as traffic congestion and environmental pollution. Private cars trajectory data and perceiving their information provide a promising solution to these problems. However, the collection of large-scale trajectory data for private cars with high accuracy and reliability is still delicate tasks in urban environments. In this paper, we propose a low-cost and user-friendly implementation method for achieving large-scale private cars trajectory data acquisition via designing lightweight GPS module and On Board Diagnostics (OBD) reader. To ensure reliable trajectory data acquisition via GPS/OBD integration, we propose an ensemble learning based Gauss Process Regression (GPR) method so as to cope with the non-linearity, non-stationarity and incremental training problems during trajectory collection. We design a classification-type loss (CTL) function and build a regression to classification (R2C) method with Learn++ for realizing ensemble learning. The proposed approach implements incremental learning when new trajectory data arrives and is able to resolve the concept drifting problem. Experiments in real-world urban environment have demonstrated the effectiveness and reliability of the proposed method, it achieves better trajectory prediction performance than the comparative methods under various road conditions in GPS-denied areas.

Integration of Second-generation On-board Diagnostics Data via Deep Learning to Develop Eco-driving Analysis System Applicable to Large and Small Cars

Integration of Second-generation On-board Diagnostics Data via Deep Learning to Develop Eco-driving Analysis System Applicable to Large and Small Cars

Study on Determination of Excessive Emissions of Heavy Diesel Trucks Based on OBD Data Repaired

It has been recognized that emission control for heavy diesel trucks should be given priority, as a massive amount of pollutants (e.g., NOx) are emitted from heavy diesel trucks. Although pollutants can be filtered to a considerable extent by after-treatment devices equipment, emissions can still exceed the designated standards when after-treatment devices function improperly. To timely identify excessive emissions, we propose a general and systematic framework, including a data quality assessment and a data repairing and excessive emission determination process, based on the data sensed from the on-board diagnostics (OBD) monitoring system. To overcome the adverse effects of poor data quality, a set of approaches have been developed for the different statuses of data quality. When all variables contain missing or abnormal values, data repairing algorithms can be employed to improve data quality. Two strategies have been developed for the situation where only the NOx data is problematic. One is to improve data quality by using the other variables before identifying excessive emissions, and the other is to directly predict whether the emissions exceed recommendations by using other variables without the data quality problem. To reduce the impact of noise and extreme values, three methods based on the moving average principle have been developed to generate an aggregated emission level for the determination of excessive emissions. In the experimental study, we employed a number of machine learning algorithms to achieve data repairing and prediction. The support vector machine (SVM) algorithm slightly outperforms the random forests (RF) and gradient boosting decision tree (GBDT) in the prediction of the excessive emission possibility in terms of prediction accuracy. The experimental results indicate that the most accurate data repairing can be achieved by probabilistic principal component analysis (PPCA), as compared to non-negative matrix factorization (NNMF) and k-nearest neighbor (KNN). However, the proposed approach does not restrict other algorithms from achieving the functions of data repairing and prediction.

Characteristics of Resistive PM Sensors for Onboard Diagnostics of Diesel Particulate Filter Failure.

In accordance with the recently reinforced exhaust regulations and onboard diagnostics regulations, it is essential to adopt diesel particulate filter systems in diesel vehicles; a sensor that directly measures particulate matter (PM) in exhaust gas is installed to precisely monitor diesel particulate filter (DPF) failure. Because the reduction of particulate matter in the diesel particulate filter system is greatly influenced by the physical wall structure of the substrate, the presence or absence of damage to the substrate wall (cracks or local melting, etc.) determines the reliability of normal DPF operation. Therefore, an onboard diagnostics sensor for particle matter is being developed with a focus on monitoring damage to the DPF wall. In this study, as a sensor for determining damage to the substrate wall, an accumulation-type sensor whose resistance changes as soot particles are deposited between two electrodes was fabricated. The sensor characteristics were investigated by changing the gap between the sensor electrodes, sensor cap shape, and electrode bias voltage to improve resistive soot sensor sensitivity and response. From the signal characteristics of various sensor configurations, a combination sensor with improved signal stability and response time is manufactured, and they were compared with the characteristics of commercially available sensors in the engine-simulated NEDC mode in terms of the degree of DPF crack. As a result of transient mode, PM monitoring cycle was improved by 1.2~1.5 times during the same vehicle driving time compared to the existing commercial sensor.

A Review of IoT Technology for the Connected Autonomous Vehicles Ecosystem

Connected Autonomous Vehicles Ecosystem (CAVE) is fast becoming a reality. This paper reviews various architectures that leverage IoT technology to progressively create a system of interconnected Autonomous Vehicles (AV) that is effective, robust and secure. Accurate and fast perception of the external environment is key to movement of AVs in CAVE. IoT technology through sensors and internetworking can help an AV to do reliable perception. State of the art methods for perception of AVs in CAVE are reviewed. Self-awareness is another important aspect. IoT and Artificial Intelligence (AI) can come together to assist an AV in self-diagnosing anything abnormal. One such use case is presented for onboard diagnostics in an AV. Accurate AI is key to making effective strategies and to take control actions with error feedbacks that carefully maneuver an AV from its source to destination. The transition from legacy vehicles to CAVE requires AI methods that combat threats and have effective overall movement in mixed mode traffic systems (MMTS). Strategies for effective movement in MMTS are presented.
 HIGHLIGHTS
 
 Role of IoT technology is identified in different areas of Connected Autonomous Vehicle Ecosystem (CAVE)
 A comparison of different technologies in various domains of CAVE is presented and their pros and cons are discussed
 Use case for onboard diagnostics of an autonomous vehicle in CAVE is presented to emphasize the role of IoT Technology in [automated] preventive maintenance
 Key studies are reviewed that emphasize important issues like robust architectures for CAVE, threats in CAVE and reliable functioning in mixed mode traffic systems
 
 GRAPHICAL ABSTRACT

Logic Solver Diagnostics in Safety Instrumented Systems for Oil and Gas Applications

A safety instrumented system (SIS) is a complex unit composed of a set of hardware and software controls which are expressly used in critical process systems. A SIS should be specifically designed to obtain the failsafe state of the monitored plant or maintain safety of the procedure or a process when unacceptable or dangerous conditions occur. This paper focuses on condition monitoring and different diagnostic solutions used in safety instrumented systems, such as limit alarm trips, on-board diagnostics, and logic solver diagnostics. A case study consisting of the design of a safety loop using standard IEC 61508 for a complex safety instrumented system in the oil and gas field is presented in the paper using a diagnostics-oriented approach. The presented methodology aims at reaching the optimal tradeoff between IEC 61508 and the market requirements focusing on the best technological solutions to optimize diagnostics and safety and minimize the system’s response time in case of failure. The results of the application emphasize the importance of an accurate diagnostic strategy on safety instrumented systems for oil and gas plants.

Mass estimation of ground vehicles based on longitudinal dynamics using loosely coupled integrated navigation system and CAN-bus data with model parameter estimation

This work considers real-time estimation of the mass of ground vehicles equipped with on-board diagnostics (OBD-II) capabilities. The mass is estimated with a recursive least squares filter, based on vehicle longitudinal dynamics using data from a 6-axis inertial measurement unit (IMU), OBD-II data, as well as global navigation satellite system (GNSS) positions. Only basic OBD-II parameters from the controller area network (CAN-bus) are used, including vehicle speed, engine speed, and engine load. This ensures a broad applicability of the presented work. A loosely coupled fusion of IMU data, OBD-II vehicle speed, and GNSS positions is used for vehicle state estimation. The road grade is estimated from the vehicle pitch angle, current gear is predicted from the ratio of engine and vehicle speeds, and engine torque is estimated from a linear regression with engine speed and load as parameters. The regression is carried out using engine torque estimated from the vehicle longitudinal dynamics, assuming a known vehicle mass. Model parameters are estimated from the engine torque regression, utilizing a linear correlation between the engine torque and OBD-II engine load. A method for automatic estimation of vehicle gearing ratio values from drive data is also proposed. The result is a two-phase approach with an initial training phase for parameter estimation, followed by an operational phase where the vehicle mass can be estimated. The methods are validated using data from a modern car. Parameters and regression coefficients are first estimated from a single test drive. Vehicle mass is then estimated using data from 18 drives on an 85 km test route, comprising of more than 1600 km of driving under different condition, including varying vehicle loads, tire pressures, and window openings. Using the fitted model parameters, the model is generally able to estimate masses within ± 5% of the actual. The change in tire pressures and windows openings do not show significant effects on the estimated masses. Ambient wind during the test drives appears to present a significant source of uncertainty in the estimates at higher speeds. A method for compensation of the ambient wind should be investigated.

Decelerating and recovery techniques for hypervelocity projectiles

Recovering the projectile, target debris and assessing damage effects are important in evaluating the performance of hypervelocity systems and in developing protective countermeasures. We report the results of experiments and simulations that have been conducted to assess the viability of decelerating and recovering hypervelocity projectiles with masses >1.5 kg and velocities >1.7 km/s. The objectives were to retain projectile integrity and retrieve data from onboard diagnostics. This required the minimization of impact loads and frictional heat generation with respect to the selection of deceleration and recovery materials. According to the requirements of decelerating and recovering the hypervelocity projectile, an anti-flaming expandable polyethylene foam, an anti-flaming expandable polystyrene foam, an anti-flaming sponge and an anti-flaming expandable polyurethane foam were chosen as the decelerating materials in the verification test while paraffin was chosen as the recovery material. Based on their damage characteristics under impact and the cost, the anti-flaming sponge and the paraffin were selected as the decelerating material and recovery material respectively. Deceleration effects on the moving projectile were obtained by numerical simulation. The device was designed and installed with anti-flaming sponge and paraffin, which successfully recovered the projectile with a mass of 1.89 kg at a velocity of 1.90 km/s. The projectile velocity profile was successfully obtained in the deceleration and recovery materials.

Study of Real-road Nitrogen Oxide Emissions of Non-road Vehicles

The exhaust gas pollutants of the non-road vehicles are harmful to the environment. Many non-road vehicles meet the requirements of the regulations in the laboratory. However, the real-road emissions of such vehicles are sometimes higher. Measuring the real-road emissions of non-road vehicles is very important. The real-road emissions are measured by on-Board Diagnostics (OBD), but there are some problems in the data stability of OBD. The NOx emissions of a bulldozer (a type of China IV non-road vehicle) based on both portable emission measurement system (PEMS) and OBD are studied in this article. Experiments contained three working processes: idle, driving, and operating. The nitrogen oxide (NOx) emissions during operating were highest. The NOx emission characteristics of the bulldozer from PEMS and OBD have the similar variation trends. But there are still some differences, including the NOx emission value and response time. The measurement principles and different sampling points between PEMS and OBD are the main factors. An effective data processing method is introduced to reduce the differences of between the data from PEMS and OBD. Briefly, the NOx emissions of the OBD and PEMS were highly consistent. The OBD is reliable and can be widely used in non-road vehicles.

Data-Driven Models Support a Vision for Over-the-Air Vehicle Emission Inspections

Emissions inspection and maintenance (I/M) programs for light-duty motor vehicles manage ambient air quality by enforcing emissions standards and requiring non-compliant vehicles to be repaired or retired. I/M programs in the United States typically identify over-emitters through on-board diagnostics (OBD) systems and vehicles’ proprietary firmware (i.e., indirect tests), rather than through physical measurements of exhaust gases (i.e., tailpipe tests). Analyzing data from Colorado’s I/M program, this study finds the OBD test to have an accuracy of 87%, but a false pass rate of 50%, when predicting the result of a corresponding tailpipe test. As an alternative, transparent data-driven models—using logistic regression and gradient boosting machines—to consistently identify over-emitting vehicles are proposed. These models were up to 24% more accurate, or 85% more sensitive than the current OBD test in a stratified data sample. A key benefit of transparent statistical models—jurisdictions’ ability to tune the test methods to best suit program needs—is also assessed. Finally, this study shows how these results support a vision for cloud-based, selective I/M programs where statistical models are applied to OBD data—collected over-the-air from vehicles—to identify and require additional inspection for only the most probable over-emitters.