The following is an additional summary of the report submitted by Mitsubishi Motors Corporation (MMC) to Ministry of Land, Infrastructure, Transport and Tourism (MLIT) today, pursuant to instructions received from MLIT on April 20 to investigate improper conduct in fuel consumption testing of vehicles manufactured by MMC. We express deep apologies to all of our customers and stakeholders for this issue
- Status of investigation into the four mini-car models (eK Wagon, eK Space, Dayz and Dayz Roox)
- Outline of improper conductThe fuel consumption targets for the fuel-economy grades of the 2014 eK Wagon and Dayz (submitted for certification in February 2013) were raised on five different occasions in consideration of competitors’ fuel consumption rates and other factors. Estimated running resistance values were used in the fuel consumption measurement to confirm whether the fuel consumption targets were reached, where ordinarily measured values from an actual test car should have been used. Furthermore, the fuel consumption figures reported and finalized at various internal meetings were based on running resistance values that had been arbitrarily altered.The running resistance values used for other grades of the same models (the standard, turbo and 4WD grades), for the eK Space and Dayz Roox models that were released later, and for those four models in subsequent model years, were all calculated on paper based on the untested running resistance value used for the 2014 eK Wagon and Dayz.
- Background to the improper conduct
- This project was considered as an important part of the joint venture with Nissan Motors.
- One of the project’s development goals was to achieve best-in-class fuel consumption in order to compete with other models in the mini-car market, which is highly competitive with respect to fuel consumption.
- The Performance Testing Department had long been engaging in setting and achieving fuel consumption targets, which was outside their authority as laid out in MMC’s “Detailed Rules of Company Organization”. Setting and achieving targets for vehicle specifications, including fuel consumption, was officially supposed to be the responsibility of the Product Development Project.
- Furthermore, part of the work of setting and achieving fuel consumption data had been wholly outsourced to Mitsubishi Automotive Engineering Co., Ltd. (a wholly-owned subsidiary of MMC; the “Subsidiary”) via the Performance Testing Department.
- Causes of the improper conductIn addition to the circumstances given above, the following possible causes of the improper conduct were put forward as a result of the interview investigation.
- The General Manager of the Performance Testing Department (the “Performance Testing GM”) and the other manager in the Performance Testing Department (the “Performance Testing Manager”) were aware of the difficulty of achieving the fuel consumptiontargets, but nevertheless completely outsourced the achievement of those fuel consumption targets to the Subsidiary, and failed to supervise the process unless there were queries or reports from the managers of the Subsidiary. Moreover, they passed on the results reported by the Subsidiary to the Final Design Quality Control Meeting and other meetings without first confirming those results themselves, which it was their duty to do.The Product Executive (the “PX”) and the Development Project Manager (the “Development PM”) likewise failed to confirm in detail the running resistance values, and merely accepted the fuel consumption values as reported.
- The manager of the Subsidiary (the “Subsidiary Manager”) believed that if he measured the running resistance in Thailand as initially planned, he would see running resistance values in line with the running resistance values estimated on paper based on past models. The Performance Testing GM and Performance Testing Manager instructed the Subsidiary Manager, before measuring the running resistance in Thailand, to collect a lot of “good” data (i.e., low fuel consumption values) using the legally-required “coasting test”. The understanding was that if the values were measured using the proper “coasting test”, there was no internal rule at MMC saying that the median values had to be used.
- However, the Subsidiary Manager performed the measurements using the “high-speed coasting test”, and he was ultimately unable to reproduce the estimated values used so far using the median measured values. The Subsidiary Manager used inappropriately low values from the data to produce running resistance values even lower than those estimated on paper, and presented these data to the Performance Testing Manager. The Performance Testing Manager checked the running resistance data that they received, and was aware that it was measured not using the legally-stipulated “coasting test” but the “high-speed coasting test.” The “high-speed coasting test” has a large margin of error and they considered it inappropriate to take the lower values, but there was no time to redo the tests, so they approved the data provided by the Subsidiary Manager. The Subsidiary Manager has no experience in testing using the legally-stipulated “coasting test”.
- The Subsidiary Manager made a statement to the effect that in order to win against competitors’ vehicles and in the context of management’s high expectations of improved fuel consumption, that he felt that the DevelopmentPM required him to commit to meeting the fuel consumption targets. Also, he felt that the Corporate General Manager of the Development Engineering Office and PX required him to commit to improving fuel consumption at all costs. He further stated that with the limited resources available it was impossible to allocate enough time for proper testing after the test car was sent to them.
- We found issues in our administration and process such as that the setting and achieving fuel consumption had been entirely performed within the Performance Testing Department with insufficient outside supervision, and that the risk was not highlighted at any point in our development process, such that we were unable to rectify the problem.
- Status of investigation into other models currently sold in Japan
- The “high-speed coasting test” was used in measuring the running resistance of all models other than Mirage, Outlander PHEV, and Delica D:5 diesel version.
- When MMC prepared test reports required by law, MMC mentioned fictional information such as test date, weather, atmospheric pressure, and air temperature.
- MMC did not measure running resistance for the RVR, and instead calculated it on paper based on improperly altered data of other models.
- For the Pajero gasoline version, we used data calculated by arbitrarily combining the low running resistance and low air resistance data of a different testing car from past measurement data.
- For the Outlander PHEV, Outlander, Delica D:5, RVR, and Pajero, we performed paper calculations based on past test results and the like when we made corrections regarding weight, improved CVT, improved tires, etc.
- MMC conducted running resistance measurements using the legally required “coasting-test,” and after confirming through internal testing, we found deviation from specification value within ±3%.
- Running resistancetesting method
The background of and the reason for MMC’s using a “high-speed coasting test,” which differed from the coasting test required by the Road Trucking Vehicle Act, are as follows:
- The background of MMC’s using a “high-speed coasting test”
- In 1991, the 10 mode cycle, which was used for measuring automobile exhaust emissions and fuel consumption, was replaced with the 10-15 mode cycle,and intake manifold pressure (boost) testing for measuring running resistance was replaced with the current coasting test. At that time, MMC adopted a “high-speed coasting test,” which was based on a “coast down test” that had been used for vehicles for the United States since 1978 and did not adopt the coasting test required by the Road Trucking Vehicle Act.
- The coasting test required by the Road Trucking Vehicle Act requires the inclusion of coasting time and average coasting time in the record of the load setting. So a manual had been in use since at least January 1992 to calculate coasting time by using the running resistance measured in a “high-speed coasting test”.
- MMC compared the coasting test to the “high-speed coasting test” in January 2001 and found that the maximum difference in results was 2.3%.
- The testing manual “Running Resistance Measuring Method” was updated in February 2007 and an appendix describing testing methods for each target market was added to the manual. The appendix included the statement that “TRIAS (i.e., the coasting test) is to be used for DOM (i.e., vehicles for sale in Japan)”.
- The reason for MMC’s using a “high-speed coasting test”
With respect to the reason why MMC adopted, in 1991, a “high-speed coasting test” which differed from the coasting test required by the Road Trucking Vehicle Act, MMC interviewed the persons in charge at the time, including retired employees, but no clear-cut answer was obtained in this regard. However, MMC assumes that the reason why it continued to use a “high-speed coasting test” even though there were several chances to replace it with the coasting test is that previous test result aforementioned 3 (1) (iii) gave only 2.3% at the maximum.
- The background of MMC’s using a “high-speed coasting test”
- Certification process for categories other than emission and fuel consumptionMMC has investigated the status of the four mini-cars and nine other models currently sold in Japan regarding 7 tests ((1) through (7) below) submitted by automakers to MLIT during the certification process. Regarding “(1) MLIT measured the running resistance of the four mini-cars, while MMC re-measured that of the nine other models using the proper coasting test. Both test confirmed that there are no problems with the results or with their technical underpinnings. Regarding “(7) Brake test (electronic connecting device)” below, MMC does not use this test.
Regarding other 5 tests (from (2) to (6) below), MMC has confirmed that there are no problems with the submitted data or with their technical underpinnings. Regarding “(2) Emission and fuel consumption test” below, since MMC has confirmed that the correction coefficient for periodical regeneration control submitted to MLIT is the same as test data, and that the impact of the correction coefficient for battery on emissions and fuel consumption is 1% or less based on test results and analysis, we considered it to bewithin the margin of error and did not report this to MLIT.Regarding “(3) Emission and fuel consumption test” below, MMC has confirmed that the emissions correction coefficient pursuant to battery charge level which was submitted to MLIT is the same as the data measured by MMC.
(1) [Emission and fuel consumption test] Running resistance value of cars (2) [Emission and fuel consumption test] Emissions correction coefficients for equipment that functions periodically, such as the DPF (Diesel Particulate Filter) in diesel cars (3) [Emission and fuel consumption test] Emissions correction coefficients pursuant to battery charge level for hybrid cars and the like (4) [Confirmation of solidness of car frame] Strength of car frame of trucks and the like (5) [Brake test] Simulation results regarding antiskid brake system of variant models (6) [Brake test] Location of the center of gravity which is necessary for the confirmation of ABS performance (7) [Brake test] Values which are necessary for confirming the performance of electronically controlled brakes
- Involvement and awareness of the management with respect to this conductNo direct instructions to carry out the improper conduct were given by management; moreover an internal system for managing the product development process was established. However the management had an insufficient understanding of the actual operational practices of the product development section. We also acknowledge that the high expectations for a shorter product development process and better fuel consumption have created an environment which could result in improper conduct, that our internal management system has not been adequately organized, and that education, training and internal supervision of the managers have not been thoroughly conducted.In addition to the above, although it should have been our first priority to carry out our business in compliance with all the applicable laws and regulations, in reality we have not done this practice for a long time. We are deeply concerned by this revelation.Although MMC has been working on various reforms to prevent the recurrence of any product quality issues under the slogans of “Compliance First,” “Customer First” and “Safety First” which were adopted based on the product quality issues in 2000 and 2004 as well as the corporate program entitled the “Customer First Program” established after the issue in 2012, such slogans and program have not been fully recognized throughout our organization and by all employees. We are taking this situation seriously, and we intend to carry on our internal reform with unwavering resolve, including making drastic revisions to the existing organizational structure and the business operation process and enhancing our internal education.
- Future Course of Action
- We are currently conducting an investigation into the models sold in Japan in the last 10 years other than the models currently sold in Japan (i.e., models no longer sold in Japan) and we intend to inform you of the result of that investigation separately.
- On April 25, 2016, we set up a special investigation committee composed solely of independent outside experts for the purpose of thorough investigation of this improper conduct and we have delegated the task to carry out an objective investigation based on an independent standpoint outside of MMC. Once we receive the reports and opinion of the special investigation committee, we intend to make an appropriate follow-up plan and inform you separately without delay.