Lean Six Sigma applied to process performance and improvement model for the development of electric scooter water-cooling green motor assembly

Abstract In response to the environmental issues triggered by global warming, worldwide companies gradually put the factor of carbon emission into the process of product lifecycle, developing green technology or adopting cleaner production aimed at sustainable development. Lean Six Sigma has advantages of cutting waste and facilitating process improvements as well as system analysis, helping enterprises create the overall business benefits in the value chain. Used in the renewable energy industry, it can promote the triple bottom line (TBL), the performance of sustainable production for corporate profit, social responsibility and environmental responsibility. Therefore, this work took the process performance of the electric scooter water-cooling green motor manufactured in Taiwan with the world’s highest density of scooters as a case study. The developed performance evaluation and improvement model for manufacturing scheduling and process quality achieved the goal of economic benefits of enhancing process quality performance by shortening manufacturing scheduling and reducing process variations with Lean Six Sigma. Besides, they could respond to the policy of energy saving and carbon reduction – replacing the traditional scooters of high-carbon emissions with the electric scooters of low emissions. Furthermore, they could bring enterprises into harmony with economic benefits, ecological benefits and social benefits.


Introduction
According to the research done by Folster and Nystrom (2010), Dejan et al. (2013) and Almer and Winkler (2017), the awareness of environmental protection has been continuously rising from the Kyoto Protocol of 1997 to the present 21st century. The Paris Climate Change Conference 2015 is a milestone for the global climate governance, encouraging the low-carbon energy transition and urging the globe to march towards the target of green growth and zero-carbon energy. Based on the study conducted by the Industrial Development Bureau of the Ministry of Economic Affairs in Taiwan, currently, the carbon emissions of electric scooters in the area of Taiwan are less than 50% of those of fuel-powered scooters. Besides, the research was done by Lambert (2017) and Staffel et al. (2017) also shows that the development of green energy can help the electric scooters' carbon emissions drop further. Moreover, Taiwan and Southeast Asian countries have much higher populations using scooters as a way of transportation than any other places in the world. Thus, to cope with the promotion and development of green electric scooters, the Taiwanese government actively promotes the development of the green-energy electric scooter industry, in order to reduce the fuel-powered scooters' impact ratio to air pollution.
In addition, according to Chen and Lyu (2009), Jacobs, Swink, and Linderman (2015), Garza-Reyes et al. (2016), Thomas et al. (2016), Cherrafi et al. (2017) and Bellisario and Pavlov (2018), Lean Six Sigma is a combination of lean production and Six Sigma thoroughly eradicating non-valueadded waste and shortening the time of process. Also, it can help to reduce the variation in the manufacturing process as well as the product defect rate (Srinivasan et al. 2016). Currently, a large number of researchers, such as Kumar et al. (2006), Andrew et al. (2008), Lee and Wei (2010), Wang and Chen (2010), Hardeman andGoethals (2011), Jacobs, Swink, andLinderman (2015), Thomas et al. (2016) and so on, apply Lean Six Sigma to the industries and receive remarkable outcomes. Besides, according to Antony, Rodgers, and Cudney (2017) Trzeciak et al. (2018) (Medical Services) and Yadav, Seth, and Desai (2018) (manufacturing industry), the most of literature related to Lean Six Sigma focused on the financial and economic performance, whereas only a few literatures emphasized the applications to promoting 3TBL/sustainable production or sustainable improvement. When producing the electric scooter water-cooling motors, there must be high process accuracy and timeliness. At the same time, product quality must be controlled through continuous improvement and immediate feedback, in order to achieve the goals of 3TBL/sustainable production and sustainable improvement. Owing to this, this work selects the systematic method, Lean Six Sigma, to reach the goal.
As stated above, Lean Six Sigma has advantages of cutting waste and facilitating process improvements as well as system analysis, helping enterprises create the overall business benefits in the value chain. Used in the renewable energy industry, it can urge enterprises to perform well in the TBL sustainable production. In addition, in response to the environmental issues triggered by global warming, numerous companies put the factor of carbon emission into the consideration of production, developing green technology or adopting cleaner production aimed at sustainable development. Therefore, this work took the 'electric scooter water-cooling green motor' manufactured in Taiwan, a region with the world's highest density of scooters (in 2017, the number of scooters was 806.56/km 2 ), as a case study. The green motor produced by the case company combining the concepts of green energy and cleaner production differs from the general motor using air to cool down the high temperature generated by operation. Instead, the green motor adopts the water-cooling and circulating method to perform a more efficient cooling way, so that it can better the output efficiency of motor operation and reduce the impact on our environment. Lean Six Sigma can help to bring enterprises into harmony with economic benefits, ecological benefits and social benefits.
The remainder of this article is organized as follows. In Section 2, this work will briefly introduce the process of electric scooter water-cooling green motor assembly and collect customer (VOC) as well as value stream mapping (VSM) to find out the non-valued parts in the green motor process. In Section 3, targeting at the non-valued parts, such as the manufacturing schedule performance and process quality level, this work will Six Sigma quality indices and manufacturing schedule performance indices. In Section 4, this work will deduce the evaluation and testing model for Six Sigma quality indices as well as manufacturing schedule performance and sort out the results to form an evaluation form for the electric scooter water-cooling motor manufacturing performance, in order to benefit the subsequent analyses and improvement tasks. In Section 5, the cause-and-effect diagram will be applied to find out the control factors which influenced the process. The Taguchi method will be used to figure out the best parameter setting to promote the process performance. Meanwhile, standard operation procedure will be adopted to assist the tool room management inside the factory to enhance the tool management and ensure the processing quality. Conclusions will be presented in Section 6.
2. Introduction to the analytic structure and process of the electric scooter water-cooling green motor assembly

Analytic structure
In this article, Lean Six Sigma is used as an analytic method, and the DMAIC procedure of Six Sigma is used as a framework, to develop a process performance evaluation and improvement model for an electric scooter water-cooled green motor assembly. First, the voice of customer is paired up with the value stream mapping to find out the valueadded activities in the process, in order to find out the problems that need to be improved. Next, this work explores the important processes of the motor assembly and summarizes the important quality characteristics. Manufacturing schedule performance indices are proposed based on important processes, and Six Sigma evaluation indices are come up with based on important quality characteristics to evaluate manufacturing schedule performance and process quality standards. This work then deduces the confidence intervals corresponding to Six Sigma quality indicators and manufacturing schedule performance indicators to evaluate and verify the performance indicators, in order to review the process characteristics and quality characteristics that need to be improved. The cause-and-effect diagram and the Taguchi method are used to analyze the control factors which influenced the process and figured out the best parameter setting, aiming to promote the process performance as well as assist the tool room management inside the factory. Last, the control chart is developed to continuously monitor the quality of the process. Lean Six Sigma can not only shorten the manufacturing time but also reduce the process variation to create business value and then the goal of 3TBL/ sustainable production and sustainable improvement can be achieved

Introduction to manufacturing process
As the above-mentioned, the water-cooling green motor assembly of the electric scooter manufactured by the casestudy company is more complicated than and different from that of the commonly sold electric scooter in the market which uses air impact to cool down the high temperature resulted from the motor's operation. The water-cooling method is adopted to conduct a more efficient way of lowering the temperature to enhance the output efficacy of the motor's operation. The water-cooling flow path created by the case-study company is to enclose the entire motor to stabilize its working temperature, so the structure of the motor consists of two major partsinner water case and outer water case, called as the motor assembly in this article. Thus, the motor assembly is divided into two major processes: inner water case underline process and outer water case inner-hole process. After the preprocessed multiple processing procedures, the airtight machine will be sent to conduct an air tightness test. The entire process is displayed in Figure 1.
In the light of the electric scooter water-cooling green motor assembly, this work adopts Lean Six Sigma as an improvement method, which can not only reduce waste but also create the overall business benefits in the value chain. If the voice of customer (VOC) can take several factors, such as environmental protection, green energy design, cleaner production and reducing carbon emissions, into consideration, it can help enterprises move towards the goals of economic benefits, ecological benefits and social benefits into harmony. This work presents the voice of customer (VOC) as shown in Table 1. Therefore, in the definition stage of Lean Six Sigma, we discover the customers' problems related to product quality and environmental protection and draw the current value stream mapping (VSM). Only when the nonvalue-added reasons caused by the motor assembly process are confirmed and the problem that needs to be improved first is identified, the entire production capacity and performance can largely enhance.
As mentioned above, this work uses VSM to understand the process flow of the electric scooter water-cooling green motor, able to show all the statuses of the process and also able to know all of the value directions and the required time, as displayed in Figure 2. According to the drawn VSM, the values of the required time information can be obtained, and the total cycle time of the process can be computed to judge whether the process can achieve lean.
This work, based on the previously stated VOC and VSM, respectively, lists the goals for improvement and their corresponding critical quality indices. Next, concerning whether the critical quality indices correspond to VOC, space for improvement can be rated as weight. Besides, according to these critical quality indices indicating whether there is a project taking place in the company, all the present performances of the critical quality indices are listed, and the target performance is seen as an evaluation to see whether these critical quality indices will be brought into the range of the project, as demonstrated in Table 2. Last, the aim of getting the project is to lift the process quality level to reduce the process defect rate (estimated to lower 5% of defect rate) and the airtightness defect rate (estimated to lower 4%); meanwhile, it can boost the manufacturing efficiency to shorten the manufacturing time (estimated to lower at least 13% of time waste).
Decreasing the defect rate of air tightness is to use the green motor manufactured by the case company combining the concepts of green energy and cleaner production and employ the water-cooling and circulating method to perform a more efficient cooling approach, so that the motor operation output efficiency can enhance. Besides, the green motor can reduce carbon emissions and achieve the environmental protection of green energy products. Furthermore, Lean Six Sigma can help promote the process quality as well as reduce the defect rate of air tightness.
Hence, this article will target at these two processes to develop the manufacturing scheduling performance indices and six sigma evaluation indices. Based on the spirit of Lean  Six Sigma, the evaluations and analyses of the manufacturing scheduling performance indices and process quality level are conducted as well. Thus, this work will also aim for the improvement of the tool room management.

Manufacturing performance indices
As the above-mentioned, the electric water-cooling green motor assembly includes the inner water case underlining process and the outer water case inner-hole process, as shown in Figure 1. There are five quality characteristics of the Normal The Better (NTB) in the inner water case underlining process, while there are four quality characteristics of NTB in the outer water case inner-hole process. The tolerance specifications of these nine quality characteristics are displayed in Table 3. This work proposes the evaluation model of two important indicators which are manufacturing quality level and manufacturing schedule performance. In addition to reaching the goal of promoting the economic benefits of process quality performance using Lean Six Sigma to shorten the manufacturing schedule as well as process variations, this case study develops green technology and adopts the method of cleaner production in response to the policy of energy saving and carbon reduction, in order to help enterprises move forward to meet the goal of sustainable development concerning economic benefits, ecological benefits and social benefits.
In regard to quality, this work will employ the concept of Chen, Chen, and Chang (2017) to develop the evaluation model for six sigma quality indicators. As to the manufacturing schedule, based on the theories of Huang and Chen (2004) and Chen and Huang (2006), the speed of manufacturing  The time of inner water hull process is too long The time of inner water hull process 5 6 11 25% 12% The time of outer water hull process is too long The time of outer water hull process 5 6 11 27% 12% schedule will influence the manufacturing quality, so the manufacturing schedule must be moderate. Therefore, this work develops the evaluation model for the manufacturing schedule performance indicators. Next, these two manufacturing performance indicators will be discussed respectively.

Six sigma quality indices
It can be seen from Table 1 that the electric scooter watercooling green motor assembly includes two processes, the inner water case underline process and the outer water case bore process, with nine nominal-the-best quality characteristics in total. This article assumes that random variable Xab represents the normal process distribution of process a and quality characteristic b and the process mean is l ab and the process variable number is r 2 ab ab , namely X ab $Nðl ab ; r 2 ab Þ. As noted by Chang, Wang, and Chen (2014), if quality characteristic X is nominal-the-better (NTB) type, then LSL ab X ab <USL ab . Then, this article, through the change of variables, lets The specifications of these nine quality characteristics are all converted intoðLSL ab ; T ab ; USL ab Þ ¼ ðÀ1; 0; 1Þ, where LSL ab is the lower specification limit, T ab is the target and USL ab is the upper specification limit. According to the theory of Chen, Chen, and Chang (2017), the six sigma quality indices of these nine quality characteristics are defined as follows: represents the event of good products which meets the specification requirements for process a and quality characteristic b and its yield is Based on De Morgan theorem and Boole's inequality, we have Obviously, if its compliment set E C ab means the event of defective products, then Similarly, let E a ¼ \ b E ab , then E a represents the event of good products for the inner water case underline process (a ¼ I) or the outer water case bore process (a ¼ II). Based on De Morgan theorem and Boole's inequality, we have According to the above equation, this article defines the process quality indices of these two processes are as follows: Based on Equations (4) and (5), we have P b ð1 À UðQ pkab À 1:5ÞÞ ¼ 1ÀUðQ pka À1:5Þ and p a ! 2 Â U Q pka À 1:5 ð Þ À1: When the quality level of these two processes reaches k 0 r(Q pka ¼ k 0 ), then the process yield p a will be least 2 Â Uðk 0 À 1:5ÞÀ1. For example, when the process quality level reaches 4:5r, then its process yield p a will be least 2 Â Uð4:5 À 1:5ÞÀ1 ¼ 99:73%.
For the same reason, let E T ¼ \ a E a , then E T means the event of good products of the electric scooter green motor assembly. Based on De Morgan theorem and Boole's inequality, we have According to the above equation, this article defines the process quality indices of the electric scooter green motor assembly as follows: Based on Equations (6) and (7), we have P a P b ð1 À UðQ pkab À 1:5ÞÞ ¼ 1ÀUðQ pk À1:5Þ and p T ! 2 Â U Q pk À 1:5 ð Þ À1: When the quality level of the electric scooter green motor assembly reaches kr (Q pk ¼ k), then its process yield p T will be ensured to be at least 2 Â Uðk À 1:5ÞÀ1.
To sum up, the electric scooter water-cooling green motor assembly consists of the inner water case underline process (a ¼ I) and the outer water case bore process (a ¼ II), with nine important quality characteristics included. If the process quality level of the motor assembly is requested to reach kr, then the quality level of the inner water case underline process can be asked to reach k 0 I r and the quality level of the outer water case bore process can be asked to reach k 0 II r. According to the theories of Chen, Wang, and Chen (2006) and Wang et al. (2011), the quality levels of these two processes must be higher than the quality level of the motor assembly. Similarly, suppose the process quality levels of the nine important quality characteristics are asked to be equal and set as k"r, then the process quality levels of the nine important quality characteristics also must be higher than the quality levels of the two processes, that is k < k 0 II < k 0 I < k 00 , where k 0 II < k 0 I because there are five important quality characteristics in the inner water case underline process while there are four important quality characteristics in the outer water case bore process. Based on Equations (5) and (7), we have According to this result, if the process quality level of the electric scooter green motor assembly is asked to reach kr (Q pk ¼ k), then we must ask for Q pkI ¼ k 0 I , Q pkII ¼ k 0 II and Q pkab ¼ k 00 . Based on Equations (8)-(10), k, k 0 I , k 0 II and k 00 can be calculated, as listed in Table 4.

Manufacturing scheduling performance indices
Concerning manufacturing scheduling, because the electric scooter water-cooling green motor assembly consists of the inner water case underline process (a ¼ I) and the outer water case bore process (a ¼ II), this article supposes that random variable T a represents the manufacturing scheduling distribution of these two processes, a normal distribution of mean u Ta and variable r 2 Ta , namely T a $Nðu Ta ; r 2 Ta Þ. According to the theory of Chen and Huang (2006), the speed of manufacturing scheduling will affect the production quality, so manufacturing scheduling must be moderate. However, we usually ask for good manufacturing quality first and favourable manufacturing performance second. Under the premise of asking for good manufacturing quality, manufacturing scheduling definitely cannot be too fast, and it will be fine as long as the upper limit of the manufacturing scheduling is regulated. Therefore, this article lets U Ta represent the upper limit of manufacturing scheduling, so that it can modify the smaller-the-better six sigma quality indices proposed by Chen, Chen, and Chang (2017) as the manufacturing scheduling performance indices as below: Obviously, the relationship between the manufacturing scheduling performance indices and the manufacturing scheduling pass rate p Ta is as follows: In other words, there is a one-on-one mathematical relationship between the manufacturing scheduling performance indices and the manufacturing scheduling pass rate p Ta .

Interval estimations of manufacturing performance indices
As mentioned above, manufacturing performance indices include six sigma quality indices and manufacturing scheduling performance indices. Because of inevitable sampling errors in samples, the reliability of point estimates is still uncertain. If the point estimates are used to judge the quality level, that may lead to misjudgment. In view of this, many researchers, such as Kanichukattu and Luke (2013) and Chen, Chen, and Chang (2017), used the method of confidence intervals to evaluate the process quality, in order to lower down the chance of misjudgment. Consequently, this article then induces the confidence intervals of these two manufacturing indices. Meanwhile, the confidence intervals are used to conduct evaluations and tests for performance indices, and results are sorted to build a manufacturing performance evaluation chart for the electric scooter green motor assembly.

The confidence intervals of six sigma quality indices
Suppose Y ab1 ; :::; Y abj ; :::; Y abn is a set of random samples converted by variables, whose sample mean and sample standard deviation are seen as follows: X abj $ N l ab ; r 2 ab n and S ab ¼ Thus, the estimator of six sigma quality indices is shown as follows:Q Under the assumption of normality, the following random variables, n À 1 s À z a 0 i =2 ffiffiffi n p þ 1:5; (13) z a 0 i =2 is the upper a 0 i =2 quantile of Nð0; 1Þ and v 2 a 0 i =2;nÀ1 is the lower a 0 i =2 quantile of the chi-square distribution with nÀ1 degrees of freedom. Let A i ¼ fLQ piab Q piab UQ piab g, i ¼ 1,2 then PðA i Þ ¼ 1Àa i and PðA c i Þ ¼ a i . Based on De Morgan theorem and Boole's inequality, we have ) PfLQ pkab Q pkab UQ pkab g ! 1Àa, where a ¼ a 1 þ a 2 , LQ pkab ¼ MinðLQ p1ab ; LQ p2ab Þ and UQ pkab ¼ MinðUQ p1ab ; UQ p2ab Þ Obviously, ½LQ pkab ; UQ pkab is the 100ð1 À aÞ% confidence intervals of the six sigma quality index Q pkab .

The confidence intervals of Q Ta
Manufacturing scheduling sample data are resulted from the random sampling of the process. Because the electric scooter motor assembly comprises two processesthe inner water case underline process (a ¼ I) and the outer water case bore process (a ¼ II), this article selects two sets of random samples, T a1 ; T a2 ; :::; T na , a ¼ I, II, drawn from the distribution of these two processes Nðu Ta ; r 2 Ta Þ, whose sample mean and sample standard deviation are displayed as follows: T aj $N l Ta ; r 2 Ta n and S Ta ¼ Therefore, the estimator of six sigma quality indices is shown as follows:Q Similarly to six sigma quality indices, under the assumption of normality, the following random variables, Ta are mutually independently distributed as Nð0; 1Þ and v 2 nÀ1 , respectively. Let z a 0 =2 is the upper a 0 =2 quantile of Nð0; 1Þ and v 2 a 0 =2;nÀ1 is the lower a 0 =2 quantile of the chi-square distribution with nÀ1 degrees of freedom. Obviously, ½LQ Ta ; UQ Ta is the 100ð1 À aÞ% confidence intervals of Q Ta .

Manufacturing performance evaluation chart
In the previous two sections, the confidence intervals between two manufacturing scheduling performance indices and nine six sigma quality indices have been induced. Next, the confidence intervals can be applied to performance evaluations as well as tests. If the process quality level of the motor assembly is set as kr, then the process quality level required value k"r can be derived from Table 2, corresponding to the nine important quality characteristics.
1. When k 00 <LQ pkab , then it is concluded that the quality level is insufficient and needs to improve, so it is labeled as 'Bad'. 2. When LQ pkab k 00 UQ pkab , then it is concluded that the quality level meets the requirements, so it is labeled as 'Okay'. 3. When UQ pkab >k 00 , then it is concluded that the quality level exceeds the requirements, so it is labeled as 'Good'. 4. When the first item to the fifth item are all labeled as 'Okay' or 'Good', then it comes to the conclusion that the quality level of the inner water case underline process meets the requirements. As long as one item is labeled as 'Bad', then it comes to the conclusion that the quality level of the inner water case underline process does not meet the requirements and needs to improve. 5. When the sixth item to the ninth item are all labeled as 'Okay' or 'Good', then it comes to the conclusion that the quality level of the outer water case bore process meets the requirements. As long as one item is labeled as 'Bad', then it comes to the conclusion that the quality level of the outer water case bore process does not meet the requirements and needs to improve. 6. When the first item to the ninth item are all labeled as 'Okay' or 'Good', then it comes to the conclusion that the quality level of the electric scooter motor assembly meets the requirements. As long as one item is labeled as 'Bad', then it comes to the conclusion that the quality level of the electric scooter motor assembly does not meet the requirements and needs to improve.
Similarly, if the manufacturing scheduling performance of these two processes is set as Q Ta ¼ v a , then the evaluation criteria of the manufacturing scheduling performance are described as follows: 1. Whenv a <LQ Ta , then it is concluded that the manufacturing scheduling performance of the ath process is not good and needs to improve, so it is labeled as 'Bad'. 2. WhenLQ Ta v a UQ Ta , then it is concluded that the manufacturing scheduling performance of the ath process meets the requirements, so it is labeled as 'Okay'. 3. Whenv a >UQ Ta , then it is concluded that the manufacturing scheduling performance of the ath process is good, so it is labeled as 'Good'.
4. When two items are both labeled as 'Okay' or 'Good', then it comes to the conclusion that the manufacturing scheduling performance of the electric scooter motor assembly is good. As long as one item is labeled as 'Bad', then it comes to the conclusion that the manufacturing scheduling performance of the electric scooter motor assembly does not meet the requirements and needs to improve.
Subsequently, according to the above-mentioned manufacturing performance evaluation criteria of the two processes for the electric scooter motor assembly, the required information for these two evaluation criteria is collected and calculated, compacted in Table 5. Then, based on the manufacturing performance evaluation criteria of the two processes, evaluations will be made.
Accordingly, this work adopts a real case to explain the application of manufacturing performance evaluation indices. The product quality of the case-study company are requested to reach five sigma (k ¼ 5); thus, based on Equations (8)-(10), the quality level of the inner water case underline process, 5.154 (k 0 I ¼ 5:154), that of the outer water case bore process, 5.211 (k 0 II ¼ 5:211), and that of all individual important quality characteristics, 5.548 (k" ¼ 5:548), can be computed. Next, this work selects 30 sets (n ¼ 30) of sampling data, calculates the parameters related to the mean of samples ( X ) and sigma (s) and applies them to Equations (13)- (14), so that the confidence intervals of six sigma evaluation indices for the related processes can be obtained, as displayed in Table 5. Furthermore, concerning the part of manufacturing scheduling performance evaluation, based on the procedures of the flow in Section 4.2 and Equations (16)-(17), the confidence intervals of the related manufacturing scheduling performance indices can be received, as shown in Table 5.
Next, this work can be based on the manufacturing performance evaluation criteria in Section 4.3 to make judgments. Concerning the quality level of the inner water case underline process, its process level is marked 'Okay', and the process levels of its five important quality characteristics are marked as 'Okay' or 'Good', showing that the process quality level and the manufacturing scheduling both meet the requirements. In addition, as to the quality level of the outer water case bore process, its process level is marked 'Bad', and among the processes of its four important quality characteristics, the motor rotor circle diameter is marked as 'Bad' as well, indicating that the process quality level and the manufacturing scheduling both do not meet the requirements and need to improve.

Process quality analysis and improvement model
According to the above-mentioned manufacturing performance evaluation chart, it was discovered that the critical factor which led to ill performance of the electric scooter green motor assembly process was the motor rotor circle diameter, whose process capability and manufacturing scheduling both failed, so the process should be reviewed and improved. Therefore, this work will employ the characteristic cause-andeffect diagram and the Taguchi method to figure out the best parameter matching, and finally use the control chart to ensure the effect of improvement, as stated separately below.

Characteristic cause-and-effect diagram analysis
At present, there are plenty of tools for process performance analyses, such as Brain Storming which can create visualization as a start, Affinity Diagramming and Structure Tree/Tree Diagram, which can help people systemize to categorize their visualization and selections. The high-level flow chart, SIPOC (read as sye-pahk) Diagram and the Flowchart/Process Map are employed to indicate the details of importance processes. Meanwhile, the Cause-and-Effect diagram or Fishbone diagram is used to figure out the possible causes of the problems and categorize them, so that other causes seem to be linked together in the structure tree to clarify the important processes or problems.
According to Srinivasan et al. (2014), the characteristic cause-and-effect diagram proposed by Professor Kaoru Ishikawa in 1976 is often used as an important tool for problem analyses as well as widely applied to the promotion of quality activities. Thus, the characteristic cause-and-effect diagram (see Figure 3) is employed in this work to analyze the reasons for the ill process performance of motor rotor circle diameter, and it is also seen as a basis to find out all factors which lead to the ill process performance of motor rotor circle diameter.
After discussing with the factory engineers and experts, this work discovered that the critical factor (see Figure 3) which caused the ill process performance of motor rotor circle diameter mainly resulted from the problem of cutting conditions in the process capability, and other important influential factors were cutting depth, cutting speed, amount of feed, and rotational speed. This work employed the Taguchi method commonly seen in the industry to figure out the best parameters to improve the accuracy of the motor rotor circle diameter process and then to enhance the entire yield and quality of the motor process. In addition, the reason that caused ill manufacturing scheduling was the tool room management. When the cutting condition in the process of the motor rotor's circle diameter is not good or there is something wrong with the cutting tool room management, it will affect the green motor operation performance and even lead to ill air-tightness of the outer water case. Thus, the effect of the motor water-cooling system will be influenced, and the problem of carbon emission will arise to affect our environment, so that the performance achievement of the environmental and social responsibilities in TBL will decline. Ross (1988), Taguchi (1991) and Chen, Wang, and Chen (2006) stated that the Taguchi method proposed by a Japanese doctor named Taguchi Genichi in the 1950s determined the best parameter with an empirical method. Besides, product design and quality robustness are provided with the lowest cost and the minimum experimental times within a short period of time. Hence, to receive the best parameter settings, this work used the Taguchi method to carry out analyses and improvement of process quality.

Taguchi method
After discussing with the experts and factory engineers, it was found that cutting depth, cutting speed, amount of feed and rotational speed were the core control factors which affected the imprecision of motor rotor circle diameter process. Additionally, each control factor has three levels listed in the control factor level list of motor rotor circle diameter process, as displayed in Table 6, and the Taguchi method is adopted to find out the best parameters for four factors and three levels, with the degree of freedom -8. Therefore, this work adopts the L 9 ð3 4 Þ orthogonal array to repeat experiments three times. According to Taguchi (1987Taguchi ( , 1991, the L 9 ð3 4 Þ orthogonal array is an orthogonal array highly recommended by Dr. Taguchi and its experimental results are also presented in Table 7. According to Taguchi (1987Taguchi ( , 1991, S/N is the short form of signal to noise. If S/N is larger, it means signal is larger than noise, so that it will be easy to judge whether the received result is signal or noise. As a result, when the ratio of S/N of the motor rotor circle diameter process for the green electric scooter motor is larger, the precision is better. In addition, the motor rotor circle diameter for the green electric scooter motor is of the normal quality characteristic in this work. Its equation is listed as follows:  Through the L 9 ð3 4 Þ orthogonal array experiments of Table 6 and Equation (18), the ratio of S/N for each experiment can be received. When the ratio of S/N is larger, it means the precision is better and also reveals that the quality of the green electric scooter motor's motor rotor circle diameter process is better. Finally, this work adopts the S/N ratios listed in Table 6 and receives the mean of S/N ratios of each level in the four factors. Then, these data will be drawn in Figure 4 to obtain the best parameter settings: A. cutting speed (0.25 mm), B. cutting speed (280 m/min), C. amount of feed (0.15 mm/rpm) and D. rotational speed. In the future, these best parameter settings will be regarded as the standards of settings, to maintain the stability of process quality and then to boost the process yield rate.

Improvement of the tool room management
The production model of the case-study manufacturer is the model of small-volume large-variety production. As to the applied cutting tools, there were excessive numbers of hilts and shanks, so that caused some difficulty in the tool room management. Good tool room management can not only reduce the waste but also cut the cost. Therefore, the tool room management is an important factor in maintaining the process yield. After discussing the factory's management and on-site engineers and observing the ill factors of the tool room management shown in the cause-and-effect diagram (see Figure 5), it was learned that the case-study manufacturer's lack of a set of regulation flow related to the tool room management was the factor which led to bad tool room management. Thus, this work based on the current situation improves and regulates the SOP of tool management in the control stage for better management.

5.4.
X 2R chart of control verifying the improvement outcomes After the motor rotor circle diameter process parameters of the electric scooter green motor and tool room management are improved, this work extracts 20 sets of sampling data in the static process which has been reset. Each set has five samples, and the Minitab 16 software is used to draw the Xbar-R Chart of control, as seen in Figure 6. It is obviously seen that both the accuracy and precision of these 20 sets of samples are within the control range of the new process, indicating that the process is static, so it can be used as a  monitoring tool for the motor rotor circle diameter process in the future.
To sum up, this work applies Lean Six Sigma to the industry related to green energy. The process of analysis, improvement and control can enhance the product quality, decrease the rates of process scrap and Reduce, increase the rates of product availability and maintenance and Reuse and raise the rate of Recycle for all spare parts. At the same time, owing to the increase of lifetime, the time of Recovery is postponed. Consequently, Lean Six Sigma not only can reach the goal of economic benefits of promoting the process quality performance but also can respond to the policy of energy saving and carbon reduction and facilitate the triple bottom line (TBL), in order to help enterprises head to the sustainable development of ecological and social benefits.

Conclusions
With the continually rapid growth of research, development and manufacturing technology, humans' consuming habits are apparently changed and influenced, and modern consumers tend to emphasize quality and efficiency more and more. Thus, enterprises must provide various, environmentally friendly and high-quality products to satisfy customers' demands and to maintain enterprises' sustainable development. However, attempting to reach the goal of sustainable development, enterprises must rebuild or improve by means of the internal process to create added value. Lean Six Sigma possesses the advantages of speeding up the process and conducting the systematic statistics analyses, so that it is a popular process management tool which can help enterprises create added value and reach the goal of sustainable development. Therefore, this article adopted 'the process performance of the electric scooter water-cooling motor assembly' as a case study, applied the process and idea of Lean Six Sigma, and developed a set of process performance evaluation and improvement model for the electric scooter water-cooling motor assembly, in order to reach the goals of shortening manufacturing scheduling time and lowering the process variation with Lean Six Sigma.
First, this work integrated VOC with VSM to figure out the non-value manufacturing scheduling and quality level and solve the problems immediately. Next, two important processes of the motor assembly were discussed and nine important quality characteristics were sorted out. In the light of manufacturing scheduling and process quality, manufacturing scheduling indices and six sigma quality evaluation indices were respectively presented. Then, the confidence intervals corresponding to these two types of indices were concluded and were employed to conduct evaluations and tests of performance indices. Also, a manufacturing performance evaluation chart of the electric scooter motor assembly was formed. According to the above-mentioned method, real samples and the performance evaluation chart were used to find out that the process of motor rotor circle diameter did not meet standards in both aspectsmanufacturing quality level and scheduling. The characteristic cause-and-effect diagram was employed to analyze and discover four items of control factors which can influence the process. The Taguchi method was used to find out four items of the best parameter settings: cutting depth (0.25 mm), cutting Speed (280 m/min), amount of feed (0.15 mm/rpm) and rotational speed (800 rpm), in order to promote the process quality. Meanwhile, standard operation procedure was also imported to assist the tool room management inside the factory to ensure the enhancement of tool room management and process quality. Study results showed that the defect rate of the processed product decreased 5%, the defect rate of the airtightness lowered down (decrease 4%), and then the manufacturing efficiency was lifted to shorten the manufacturing time (estimated to lower 13% of time waste). This model constructed by this work is also applicable to other industries' process performance evaluations and improvement, aiming to reach the goals of shortening the X ÀR chart of control after improvement.
manufacturing time and reducing the process variation with Lean Six Sigma. Furthermore, this work applied the method of Lean Six Sigma to the issue related to green energy and promoted the performance of 3TBL/sustainable production. Because this work is of great value and importance in terms of sustainable development of green energy, it can supplement the insufficiency of literature concerning sustainable improvement. In terms of management implications, because of the fact that the product of this work case must be manufactured with high process accuracy and timeliness, it is necessary to grasp product quality and resources through continuous improvement and immediate feedback in order to achieve the expected performance of water-cooling green motors. If Lean Six Sigma is not used, manufacturing costs may be wasted due to differences in manufacturing process and time, and product performance, so that the goal of 3TBL/sustainable production, as well as the performance needed by the product, cannot be achieved. Therefore, it is relatively important for this work to select the systematic approach, Lean Six Sigma, to reach the goal. In addition, this work has faced the following research limitations: (1) though this work applied Lean Six Sigma to the green motor, most of the restrictions focused on the issue of the process quality. This work suggests that the future researchers be able to select the cases which can accelerate the development of 3TBL/sustainable production. (2) When this work was collecting VOC, the content related to the environment and society seemed insufficient. Therefore, this work suggests that the future researchers be able to make further discussions on the VOC's impact on the environment and society, in order to obtain more complete applications and benefits of 3TBL.

Disclosure statement
No potential conflict of interest was reported by the authors. Dr. Kuen-Suan Chen is currently a Lifetime Distinguished Professor in the Department of Industrial Engineering and Management at National Chin-Yi University of Technology (NCUT), Taiwan. Dr. Chen received his Ph.D. degree in Industrial Engineering and Management from the National Chiao Tung University in 1995. His current research interests include statistical process control, quality management, process capability analysis, performance evaluation method, six sigma and service management. He has published in International Journal of Production Research, International Journal of Production Economics, Journal of Quality Technology, Quality Engineering, Quality and Reliability Engineering International, Applied Mathematical Modelling, and European Journal of Industrial Engineering.
Dr. Kim Hua Tan is a Professor of Operations and Innovation Management at Nottingham University Business School and previously Head of the Operations Management and Information Management Division. Prior to this, he was a researcher and teaching assistant at Centre for Strategy and Performance, University of Cambridge. Dr. Tan spent many years in industry, holding various executive positions before joining academia in 1999. His current research interests are lean management, innovation management, sustainability and operations strategy. He has spoken on these subjects across the globe, including China, Taiwan, Japan, Latin America, Europe and other locales. Dr. Tan has consulted many Fortune 500 companies and appointed as Our Common Future Fellow by the Volkswagen Foundation in 2009.