1. The Important Role of Online Inspection Technology in Modern Spinning Engineering In the quality inspection of textile products, fabric surface inspection is generally carried out using a 100% inspection method; however, for cotton yarn quality inspection, since most inspections involve destructive testing, only random sampling inspection can be used. Compared with weaving engineering, spinning engineering has more production units (i.e., spinning spindles). During random sampling inspection, it is very likely that samples from some spindles with quality defects cannot be collected for testing for a long time. Furthermore, the average quality of cotton yarn, which represents the overall level based on the sample level, cannot completely determine the fabric surface quality. Often, in a batch of cotton yarn with a generally good quality, a few defective packages can affect the overall fabric surface quality and usable quality. With the continuous advancement of textile technology, especially the rapid development of electronic information technology and computer technology, online inspection technology is being increasingly widely used in spinning engineering. Accelerating the development and application of online inspection technology is of great significance for strengthening process quality control in spinning production, improving and perfecting the shortcomings of traditional yarn quality inspection methods, improving the quality level of textiles, realizing the modernization of spinning production, and promoting the progress of textile technology. Online inspection refers to methods for directly detecting the characteristics of the product being inspected during the production process. Online inspection and automatic adjustment/detection technologies have been used extensively in many processes of the spinning industry. Examples include: automatic detection of bale height and automatic detection and control of cotton box storage in the cotton cleaning process; self-leveling systems and online detection of neps and impurities in the carding process; self-leveling systems in the drawing process; online tension monitoring systems for roving frames; automatic yarn breakage detection in the spinning process; and electronic yarn clearing functions in the winding process. In the spinning production process, the drawing process is the final step in controlling sliver quality, while winding is the final step in removing defects and controlling yarn quality. The most widely used and effective online inspection technologies are those for drawing frames and winding machines. This article uses typical examples such as the Vouk SH802E, Unimax drawing frames, and Savio Orion winding machines to discuss the application and practical effects of online inspection technologies in the drawing and winding processes. 2. Application of Typical Online Inspection Technology in Key Spinning Processes 2.1 Application of Online Inspection Technology in the Drawing Frame Process The online inspection technology of the SH802-E and Unimax drawing frames includes a self-leveling system, a quality inspection and control system, and a mechanical alarm system. 2.1.1 Self-Leveling System of the Drawing Frame The working principle of the USC and USG self-leveling systems on the SH802-E and Unimax drawing frames is to continuously sample the sliver fed into the drawing frame using a pair of concave and convex rollers, detecting changes in the feed amount. The displacement of the concave and convex rollers is converted into a voltage signal, which, after computer processing, controls the servo motor to change the draft ratio of the main drafting zone of the drafting system, so that the output sliver achieves the purpose of leveling. Compared with traditional non-self-leveling drawing frames, self-leveling drawing frames have the following advantages in controlling the quality of the drawn frame output sliver: (1) The change in the draft ratio in the drafting zone is synchronized with the change in the feed amount in time. The traditional method of controlling the weight of drawn frame spinning is to sample the sliver at regular intervals during each shift of the drawing frame production process, compare the detected weight with the design weight, and when the offline detected weight exceeds the design weight by a certain range, manually adjust the drafting gear to change the draft ratio so that the sliver weight matches (or is close to) the design weight. Slivers that are too heavy or too light are recycled as unqualified products. When the feed sliver changes systematically due to reasons in the previous process (such as changes in cotton blending, changes in the waste cotton process, etc.), this method can adjust the weight of the output sliver so that the yarn number is consistent. However, there is a significant lag in the adjustment process. When the feed sliver changes irregularly due to occasional factors (such as cotton roll holes, cotton web spillage, whitening, missing sliver, etc.), it is difficult to detect the changes by timed sampling. Even if such changes are accidentally detected, it is possible to generalize and easily adjust the draft ratio of the drawing frame, thereby producing more serious system unevenness. The USC and USG type open-loop self-adjusting leveling system adjusts the delay of the draft ratio change by setting the dead zone length, so as to achieve synchronization between the draft change and the feed amount change, effectively avoiding the adjustment lag and passive control of irregular changes in the traditional adjustment method. (2) The change of draft ratio in the drafting zone is synchronized with the change of feed amount. In the traditional drawing frame spinning weight control method, it is usually stipulated that when the output sliver deviates from the standard weight by more than ±1%, one crown tooth of the drawing frame should be adjusted, and when it deviates from the standard weight by more than ±2%, the sliver should be returned to the original state. Even if this artificial rule is reasonable, if the interference factors in the drafting process of roving and yarn are excluded, when the difference in the weight of the finished sliver is ±1%, the weight of the finished yarn will also have a difference of ±1%. In the drafting mechanism of a drawing frame, adjusting a single crown tooth can only change the drafting ratio by about 1%. To make even minor adjustments, one would either need to increase the number of crown teeth or the number of teeth to be drafted, which is difficult in mechanical design. When comparing the measured weight with the standard weight in offline testing, the standard weight is either the dry weight or the wet weight at a certain moisture regain, while the measured weight is the wet weight at the actual moisture regain in the production environment. The actual moisture regain is unknown during weight testing and can only be calculated after a considerable drying period. Therefore, adjusting the drawing frame weight based on the measured weight always assumes that the moisture regain at the time of testing is consistent with the moisture regain at the previous testing, which is inaccurate. In a self-leveling drawing frame, when there is a difference ΔG between the actual feed amount G and the theoretical feed amount Go, the actual total draft ratio Q1 and the theoretical draft ratio Qo follow the following leveling equation: Q1 = Qo * (1 + ΔG/Go). Theoretically, this formula shows that whenever there is a difference ΔG between the actual feed amount G and the theoretical feed amount Go on the drawing frame, there will always be a corresponding actual draft ratio Q1, and the change is a continuous linear progression. In practical applications, by adjusting the magnification of the self-leveling leveler, the responsiveness of the draft ratio to changes in the feed amount is corrected, ensuring that the relative change in the draft ratio is consistent with the relative change in the feed amount. Furthermore, the detection mechanism of the USC and USG type self-leveling levelers is a combination of mechanical and electrical components, effectively eliminating the influence of moisture regain in traditional weight adjustment methods. 2.1.2 Quality Inspection System and Quality Control System of Drawing Frame The SH802-E and Unimax drawing frames are equipped with a quality inspection system after the front roller and before the output compression roller (stepped roller). The system uses a high-precision sliver pressure sensor (FP trumpet head) to precisely test the volume change of the output sliver. The results of the sliver quality inspection are processed by a computer and displayed on the computer screen. Although the quality monitoring system does not participate in the operation of the self-adjusting system, the results of the quality monitoring can provide information so that users can keep abreast of the quality status and guide the adjustment of the working state of the self-adjusting system. The output of its online detection includes: (1) A%-CV% value: representing the current sliver weight deviation A%, the weight CV% value and the sliver CV% value of the 1-meter, 3-meter, 10-meter and 100-meter sliver segments. (2) Instantaneous sliver weight change: the instantaneous weight change of the 3-meter sliver is shown in animation form. (3) Sliver weight deviation curve: shows the variation of sliver weight deviation per minute in the previous hour, per hour in the previous day, and per day in the previous month for 1 meter, 3 meters, and the previous hour. (4) Sliver evenness CV% value histogram: shows the average variation of sliver evenness CV% value per minute in the previous hour, per hour in the previous day, and per day in the previous month. (5) Instantaneous draft ratio change: shows the change of instantaneous total draft ratio in animation form. (6) Draft ratio: shows the instantaneous total draft ratio and back zone draft ratio value. At the same time, the system also sets two types of online quality control methods: one is that when the sliver quality characteristic value exceeds a certain value, the machine alarms; the other is that when the sliver quality characteristic value exceeds a certain value, the machine automatically stops. The main settings are: (1) A%: long segment weight deviation quality alarm and stop setting. (2) A%-S: short segment weight deviation quality alarm and stop setting. (3) CV%: evenness CV% value quality alarm and stop setting. (4) CVL: Long segment weight CV% value quality alarm and stop setting. (5) Minimum speed for sliver detection: This indicates that the online detection system starts detection when this speed is reached. This value is generally 75%-95% of the normal production speed. (6) Minimum speed for sliver control: This indicates that the leveling system starts controlling the sliver when this speed is reached. This speed setting must never exceed the sliver feeding speed during normal production. 2.1.3 Application and management requirements of the online detection system for drawing frames The online detection system plays an important role in modern drawing frames. However, in practical applications, it must be reasonably set and adjusted to make its role better and ensure the excellent quality of the sliver. These adjustments mainly include the parameter adjustment of the self-adjusting leveler and the adjustment of the alarm system. (1) Adjustment of Dead Zone Length, Magnification, and Mechanical Drafting Ratio during Variety Adjustment In the USC and USG type self-regulating leveling instruments, dead zone length and magnification are two important process parameters. Dead zone length determines the response delay time of the leveling mechanism to fluctuations in the fed sliver; magnification determines the degree of response of the leveling mechanism to the feed amount. The former depends on the position of the detection point on the machine, the average fiber length and dispersion, and the inertia of the system; the latter depends on the raw material properties, density, and bulkiness. When the variety is changed daily and the production raw materials are changed, the average length, dispersion, density, and bulkiness of the raw materials all change. Therefore, the dead zone length and magnification must be adjusted accordingly. The mechanical drafting ratio should be adjusted according to the required number of the fed sliver and the output sliver. The weight drafting ratio should be calculated, and the drafting gear should be selected according to the drafting table provided in the equipment manual. With the self-regulating leveling instrument closed, the weight of the output sliver should be detected offline, and then the drafting gear should be finely adjusted to make the weight of the output sliver meet the design requirements. (2) Setting the minimum speed for sliver detection and the minimum speed for sliver control. The minimum speed for sliver detection refers to the minimum speed required for the online quality inspection system of the drawing frame to detect the output sliver. When the machine's production speed is lower than this speed, the quality inspection system will not detect the quality of the output sliver. Its commonly used selection range is 75% to 95% of the normal production speed, and it is often too low. In daily production, when the production speed of the drawing frame needs to be adjusted for some reason, especially when the production speed is reduced, this setting value must be adjusted accordingly, otherwise the quality inspection system may not work. The minimum speed for sliver control refers to the minimum speed required for the sliver to be fed when the self-leveling system is working. When the sliver feeding speed is lower than this setting value, the self-leveling system will not work. In principle, it should not exceed the sliver feeding speed during normal production as the upper limit. In daily production, it is recommended to set this value as the value obtained by dividing the minimum speed for sliver detection by the mechanical draft ratio. Similarly, when the production speed is adjusted, this speed must be reset, otherwise the self-leveling system may not work. During the production process of the drawing frame, there will always be a process of stopping and starting and restarting after changing the can. During this process, the output speed and feed speed of the drawing frame will also gradually increase from zero to normal speed. Therefore, there is more than one such sliver in each output sliver. It has not been self-leveled and has not been quality inspected when starting. The length of this sliver varies depending on the production speed and the speed limit setting, usually around 100 to 150 meters. This sliver has no quality record and is also a hidden danger affecting the quality of subsequent processes. In use, two methods are usually used to control it: one is to adjust the mechanical drafting configuration of the drawing frame once a week to make the mechanical drafting as close as possible to the actual weight drafting ratio, so as to reduce the sliver weight deviation caused by the self-leveling system not working when starting; the other is to pull off two layers of sliver from the bottom of each can when changing the roving to make it a return sliver, so as to reduce the impact on subsequent processes. (3) Correction and application of online detection quality data The online quality detection of SH802E and Unimax drawing frames is realized by sliver pressure sensors. The measurement principle of sliver evenness CV% differs from that of a capacitive evenness meter, and its reflection of sliver weight deviation differs from direct weighing using a balance. Therefore, there are differences between online and offline quality data. The sliver evenness CV% value measured online on SH802E and Unimax drawing frames is typically lower than that measured offline. By repeatedly reading the online and offline CV% values, regression analysis can be used to calculate the relationship between the two, thus indirectly reflecting the actual sliver evenness level through the online CV% value. The online CV% value can reflect differences between drawing frames and between individual workers horizontally; vertically, it can be used to compare differences between shifts and the trend of evenness changes over time, identifying patterns and taking measures to control quality. The sliver weight deviation A% displayed online is also detected by a sliver pressure sensor, converted into a signal, processed by a computer, and displayed. Due to various factors in the working environment, the working point of the quality inspection system often shifts, causing a difference between the online sliver weight deviation and the actual weight deviation of the sliver detected offline. When this difference is too large, it will cause distortion of quality control and frequent machine alarms or shutdowns. Therefore, in actual production, the output sliver weight must be corrected regularly, and the monitoring working point must be adjusted. The purpose of this adjustment is twofold: firstly, to adjust the output sliver weight to the standard state; and secondly, to automatically adjust the weight deviation displayed by the online inspection to zero when the output sliver weight is equal to the standard weight. 2.2 Application of Online Inspection Technology in the Winding Process The development of automatic winding machines is one of the landmark achievements of modern textile technology. The development of modern computer application technology has made the online inspection function of automatic winding machines increasingly perfect. The combination of Savio's Orion winding machine and Uster's UQC electronic yarn clearing system represents the world's advanced level of automatic winding. In addition to its characteristics of high speed, high quality, and energy saving in production, its intelligent and integrated online inspection function is particularly prominent. The Orion type winding machine's online inspection system consists of two parts: an electronic yarn clearing system and a production monitoring system. 2.2.1 The winding machine's electronic yarn clearing system (UQC) has the same basic function as the traditional electronic yarn clearing system: it performs online yarn inspection via a detection head, obtains and processes data related to yarn quality, removes defects that meet set conditions, and ensures yarn quality. However, it has significant improvements in performance, mainly in the types of yarn defects and their corresponding settings, yarn quality monitoring and alarms, and special monitoring functions for the electronic yarn clearing hardware. (1) Yarn Defect Types and Yarn Clearing Settings of UQC Electronic Yarn Clearer In addition to the three yarn clearing channels of conventional electronic yarn clearers—short thick place (S), long thick place (L), and thin place (T)—UQC electronic yarn clearer adds twelve new yarn clearing channels: extremely short thick place (N), starting miscount (C), production miscount (CC), dark foreign fiber (FD), light foreign fiber (FL), chain foreign fiber (MF), chain yarn defect (PC), core-spun yarn defect (CY), fancy yarn defect (K), multi-head yarn (DY), double yarn defect (U), and twisted yarn defect (J), for a total of 15 yarn clearing channels. Among them, the NS, L, T, FD, and FL channels each have six auxiliary points (H1-H6) for fine-tuning the yarn clearing settings, allowing precise adjustment of the yarn clearing curve as needed. Various yarn clearing parameters can be set by inputting the keyboard through the control box. The yarn clearing curve can be accurately displayed on the LCD screen. The yarn clearing curve can be adjusted accordingly based on the distribution of yarn defects in the yarn defect classification diagram. (2) Yarn quality monitoring and quality alarm of UQC electronic yarn clearer The UQC electronic yarn clearer detection head detects the change of yarn cross section through the capacitive detection principle and the change of yarn color through the photoelectric detection principle. After computer data processing, the yarn quality information of each spindle, each work group, and the whole machine and the quality information after winding are displayed on the screen and can be printed out as needed. (a) Display the yarn defects of each level and type, the number of quality alarms, the number of cuts, the winding length and the average yarn evenness CV% value of each work unit (each spindle, each work group, and the whole machine) in a histogram. (b) Display the total number of yarn defects of each level and type and the number per 100,000 meters in each work unit in a table form, the total number of quality alarms and cuts and the number per 100,000 meters. (c) Display the average yarn evenness CV% value and the number of common yarn defects at each level in tabular form for each work unit. (d) Statistically display the number of yarn defects in each work unit, the total number of yarn defects removed, the number per 100,000 meters, and their distribution above and below the yarn clearing curve in chart form, similar to a yarn defect grading instrument. (e) Display the total number of yarn defects in each work unit, the number per 100,000 meters, and their distribution above and below the yarn clearing curve in a scatter plot on the yarn defect grading chart, according to each level and type of yarn defect. These online detection data statistics, data processing, and online display functions of UQC can provide textile mills with quality data information at any time, enabling timely feedback on yarn defect information. This facilitates targeted adjustments to parameters, making yarn clearing settings more reasonable. Simultaneously, using the quality data statistically analyzed from these online detections, quality alarm limits for various types of yarn defects can be set. When the number of yarn defects within a specified yarn length reaches the set limit, the machine alarms and cuts the yarn, while the winding spindle stops working, recording the alarm content. The quality alarms include: extremely short thick places (N), short thick places (S), long thick places (L), long thin places (T), foreign fibers (F), starting miscount (C), production miscount (CC), chain foreign fibers (MF), and chain yarn defects (PC), etc. The parameters for these alarms include reference length and the number of set defects. UQC quality alarms also include the following: (a) Alarm setting for yarn quality unevenness CV% value: Based on the average value of yarn unevenness CV% value of a certain spindle or a certain work group, according to the quality requirements of the subsequent process, the alarm upper limit and alarm lower limit of the yarn unevenness CV% value are set respectively. When the yarn unevenness CV% value exceeds the set upper limit or falls below the set lower limit during the production process, an alarm is generated; Based on the yarn unevenness CV% value of a certain work group on the winding machine, the relative values ​​of the upper and lower limits of the unevenness alarm for a single spindle in that group are set. When the unevenness CV% value of any spindle in this group exceeds the set upper limit or falls below the set lower limit relative to the average unevenness of the group, this spindle generates a quality alarm. This ensures that the evenness value of each spindle on the winding machine remains within a certain absolute range, while also ensuring that the evenness values ​​of each spindle in each work group remain relatively stable. (b) Alarm setting for specific yarn defects: According to the definition of 23 levels of yarn defects in the yarn defect classification diagram, select yarn defects of levels 1 to 5, and set the number of times they are allowed to occur within a certain length range. For example, set the maximum number of B3 and D2 level yarn defects to be 8 and 3 respectively in 100,000 meters. When the selected yarn defect in the bobbin exceeds the set number, a quality alarm is generated. (c) Alarm setting for frequently occurring yarn defects: According to the selected frequently occurring yarn defect classification level, set the number of times neps, thick places, and thin places are allowed to occur within a certain length range. Usually, the selected ranges for neps, thick places, and thin places are +200%, +50%, and -50%, respectively. When any type of frequently occurring yarn defect in the bobbin exceeds the set number, an alarm is generated. When the UQC yarn clearer triggers a quality alarm, three types of yarn clearing actions can be selected: NONE, CUT, and BLOCK. Since there are always certain system differences between online and offline quality data, when setting the values ​​for the above quality alarms, it is necessary to use the online quality data as a benchmark, and after reading the online detection data and performing statistical analysis, gradually set the values ​​from coarse to fine. Different products and the same product at different times and under different production conditions have different quality indicators, so setting the quality alarms reasonably is a long-term and meticulous task. When the above quality alarm limit value is set to zero, it means that the detection function does not work. (3) Special monitoring of detectors by UQC electronic yarn clearer The high-efficiency production of the winding machine requires the yarn clearer to work well with the winding machine. The special monitoring function of UQC is set to ensure the close cooperation between them. Its special monitoring functions are as follows: (a) Zero point monitoring (ZPM): Ensures that the zero point of the measuring head is readjusted during the twisting of the winding machine. (b) Cutter Detection (CMT): Ensures successful yarn removal during the clearer's shearing process. An alarm will be triggered if yarn continues to move in the detection slot after the cutter cuts. (c) Yarn Jump Detection (JPM): A cutting action is triggered when the yarn leaves the detection slot for a short period due to a sudden fluctuation in winding tension caused by some reason. (d) Yarn Jump Alarm (JPA): If the yarn jumps out of the detection slot for a short period, the winding spindle will stop working after the cutter cuts, and an alarm will be triggered. (e) Drum Signal Monitoring (DSM): If the detection head does not receive a drum signal after the winding spindle is started, a technical alarm will be triggered, reminding the user to check the drum sensor. (f) Drum Entanglement Monitoring (DWM): Prevents yarn from tangling in the drum, resulting in a cutting action. (g) Drum Entanglement Alarm (DWA): The yarn is cut after tangling in the drum, and an alarm is triggered simultaneously, preventing the drum from continuing to work. The above-mentioned special monitoring functions of the UQC electronic yarn clearer make the cooperation between the clearer and the winding machine closer and the function more fully utilized, which is conducive to the high production and efficiency of the winding machine, ensuring the reliability of product quality and the normal operation of the winding machine. 2.2.2 Production monitoring system of the winding machine. The production monitoring system of the Orion automatic winding machine is a complete computer information management system with three functions: setting the working parameters of the winding machine, outputting the production data of the product, and viewing the reasons for the process alarm. It can monitor the quality of yarn and bobbin, alarm the relevant data that deviates from the quality standard, set and control the working parameters of the single spindle of the winding machine, the twister and the tension controller, etc., and monitor the doffing carriage, the blowing and suction device, etc. All online detection data can be displayed in text or charts, and can be printed out when necessary. (1) Parameters and settings of the quality control of the winding machine The quality control parameters of the Orion automatic winding machine mainly include: the setting of the maximum number of defects in the bobbin, the maximum number of knots in the bobbin and the setting of the winding speed change curve. The maximum number of defects per bobbin indicates the number of defects allowed during bobbin winding. When a bobbin reaches this set value, the spindle stops working in an alarm state, and the bobbin is replaced upon restarting. The maximum number of knots per cone indicates the maximum number of knots allowed in a single cone. Because any type of knot will eventually appear as a defect in another form in subsequent processes, when the number of knots in a cone reaches this set value, the spindle stops working in an alarm state, and the operator manually lowers the cone, sets the winding length, and restarts production. Properly setting and utilizing these two parameters can eliminate bobbins with a high number of defects, control the number of knots in the cone, and thus control the quality of the cone. Winding speed, besides being related to production efficiency, directly affects the breakage rate and the rate of increase in cone hairiness. The winding speed variation curve for a single bobbin includes average speed, acceleration, maximum speed change rate, winding deceleration start position, and the control range between the upper and lower bobbins. The starting position of winding deceleration and the control range of the upper and lower bobbins are all percentages relative to the weight or length of the entire bobbin yarn. When the production variety changes, the correct product number and bobbin weight are input, and the control system can automatically calculate the speed change curve according to the set percentage, reduce the breakage rate and maintain the consistency of the bobbin yarn hair. (2) Parameters and settings of winding machine production control. The production control parameters of Orion automatic winding machine mainly include: the functional state of the twisting cycle, the function of the winding tension controller, the working performance of the anti-overlap function and the configuration of auxiliary facilities. The functional state of the twisting cycle refers to the speed change during one twisting process, the setting of the number of times the large and small suction nozzles guide the yarn and the configuration of the working parameters of the twister. Tension control is completed by the closed-loop control system adjusting the disc tensioner. The tension sensor instantaneously detects the change value of dynamic tension during the unwinding process of the yarn and converts the displacement change of the elastic element in the sensor into an electrical signal. After being processed by the computer, the signal that needs to be adjusted is transmitted to the tensioner, and the electromagnetic pressure of the tensioner is controlled to compensate for the winding tension. Tension control parameters include tension disc pressure, tension upper and lower limit percentages, and tension increments during start-up and twisting cycles. When the online-detected winding tension exceeds the set upper limit or falls below the lower limit, the closed-loop control system automatically increases or decreases the tension disc pressure to stabilize the winding tension. Before reaching normal production speed after start-up and twisting cycles, the winding tension of a single spinning spindle will decrease due to the lower speed. Setting the tension increment at this time can reduce the tension difference in the bobbins and prevent bobbins from slipping off during subsequent processing. The anti-overlap function is achieved by computer-controlled transmission ratio between the slotted bobbin and the yarn package. The main control parameters include: yarn package diameter (a representative index representing the relationship between yarn winding length and bobbin diameter), speed increase percentage, total speed increase/deceleration time, and time allocation. All parameters can be entered on the control panel. When the production product changes, the yarn package diameter parameter is calculated using the formula: C=23000*10/metric count, and entered in the corresponding menu. Reasonable parameter settings can ensure good yarn package formation. The monitoring system of Savio Orion automatic winding machine is a good example of the application of online detection technology in modern spinning engineering, which is mainly reflected in the expansion of monitoring content and the depth of monitoring quality. In terms of control content, anti-overlap, tension, knotting, electric cleaning, and rewading treatment are centrally processed by computer to realize single-spindle control. In terms of monitoring depth, it reflects the concept of product quality as the center, from normal yarn winding to start-up after yarn breakage and replacement, from centralized setting of winding tension to automatic tension adjustment in the production process, and from reasonable setting of electric cleaning yarn clearing curve to limitation of specific yarn defects, realizing high speed and high efficiency of winding machine and improvement of yarn quality. 3. Conclusion (1) Online detection technology plays an increasingly important role in modern spinning engineering. The rapid development of electronic technology and computer application technology has promoted the depth and breadth of the application of online detection technology in spinning engineering. In-depth research and vigorous promotion of online detection technology are of far-reaching significance for improving and perfecting the shortcomings of traditional quality control methods in spinning production, stabilizing and improving the quality of cotton spinning products, and enhancing the international competitiveness of my country's textile products. (2) Among imported key equipment, online detection systems and automatic control systems account for an increasingly larger proportion of equipment investment. Only by digesting and absorbing their technologies and fully utilizing their functions can limited investment achieve maximum benefits. At the same time, it is necessary to increase investment in the development and research of online detection technologies, develop control software and sensor technologies suitable for China's national conditions, and accelerate the application of online detection technologies in domestically produced textile machinery products. (3) Textile enterprises should strengthen the technical training of equipment maintenance teams, especially in the application of mechatronics and computer technology, to adapt to the maintenance needs of new textile equipment and ensure the normal functioning of the software and hardware of new equipment. Textile enterprises should attach importance to the data application and data maintenance analysis of various online detection systems, combining offline and online detection quality data in the quality control process, giving full play to their respective characteristics, and achieving effective control of product quality through the comprehensive application of online and offline detection quality control methods.