Converging global resources; Leading as industry professional

[Technology] key elements of intelligent plant layout -MES and APS system

time:2018-09-06 Face Book:

In order to realize the intellectualization of workshop, it is necessary to collect and analyze the information of production status, equipment status, energy consumption, production quality and material consumption in real time, arrange production efficiently and reasonably, improve equipment utilization rate, improve product quality, realize production traceability, feed error prevention and improve production efficiency.

The manufacturing execution system MES and the advanced production scheduling APS system are the key to improve the intelligent level of the workshop.

The difference between them

For intelligent factory, MES is the core of intelligent factory. It integrates product data management in front-end product design and process definition stage with production data management in back-end manufacturing stage to realize closed-loop collaborative life cycle management of product design, production process and maintenance service. It contains 11 modules: 1 production plan scheduling. 2. Operation personnel management. 3, production unit allocation. 4, resource status management. 5. Product tracking management. 6. Quality management. 7. Document drawing management. 8, equipment maintenance management. 9, equipment performance analysis. 10, workshop data collection. 11, manufacturing process management.

APS is the advanced planning schedule. It should be said that APS was originally a module of MES, perhaps because the optimization of production scheduling is too important, the technical threshold is too high, only to be taken out as a separate functional software. APS should satisfy the resource constraints and balance all kinds of production resources in the production process; give the optimal production scheduling plan in different production bottleneck stages; achieve rapid scheduling and respond quickly to changes in demand.

Scheduling is sorting, that is, what to do first and what to do afterwards. But you can imagine how hundreds of large and small equipment, hundreds of people are doing all kinds of tasks at the same time, how can we achieve the optimal goal (delivery time, equipment efficiency, minimum cost, etc.) under various constraints (equipment capacity, personnel, time, venues, materials, etc.)?

Let's take a simple sorting example: Suppose the computer can process 1,000,000 sequences per second, we want to build an optimal scheduling system, nine jobs can be completed in less than a second, eleven jobs in a minute, and given 20 jobs, it takes 77,147 years to find the optimal scheduling! The actual scheduling problem involves hundreds of devices and thousands of jobs, which shows that the large-scale system scheduling problem is very complex. Of course, people will not calculate in an exhaustive way.

Over the years, co-ordinators and computer experts have been looking for a quick way to optimize large-scale systems. These algorithms have their own characteristics for some specific requirements, some of which are "fast", but the results are not optimal, some of which converge very slowly and are not practical. Even academic circles have questioned whether there is an optimal solution. Until a few years ago, an American applied mathematician (EYUAN SHI) invented a partitioned nesting (NP) algorithm to prove the generation of Markov chains, achieve global convergence, and can give the confidence interval of the optimal solution. This is a shortcut to solve the complex system optimization problem of large scale systems.

Demand for APS

APS is an enterprise management software. It has a highly intelligent scheduling function for production planning. It can make full use of the resource conditions of the enterprise and find the best scheduling results in the production process with complex multi-task conditions and many constraints. The core of APS is an optimized operation engine with the best results.

In actual production, discrete industrial enterprises (small batch, multi-variety, order changes greatly), complex tasks, resources, process flow, many constraints, and is completely a dynamic process. What the enterprise needs is to schedule an optimal schedule within a tolerable time (for example, 10 minutes). And the degree to which the schedule is optimized can be judged and quantified, and the impact on the future can be predicted (see, for example, three months from now).

APS must be highly adaptable. The actual operation of enterprises may encounter different requirements at different times. For example, sometimes the shortest delivery time is required, sometimes the best equipment utilization rate is required, sometimes the minimum inventory is required, and sometimes the emergency insertion order is required. APS must be very convenient to meet the needs of enterprises at different times according to the needs of enterprises. Please.

The human-machine interface of APS must conform to the thinking mode and scheduling habit of enterprise dispatchers. It is unacceptable for enterprise users to let people undergo extremely complex training to meet the requirements of computers.


Industry status and trends

APS has many successful applications in enterprises, especially the integration with MES modules. The scheduling problem of process industries such as steel and chemical industry is relatively simple. Therefore, the optimal scheduling is easy to implement.

Because of the complexity of scheduling problem, almost all APS systems adopt rule or heuristic algorithm in discrete manufacturing. The greatest advantage of rule method or heuristic algorithm is that it can get a feasible scheduling result quickly, but it can not guarantee the optimal solution and quantify the scheduling result. For simple processes, fewer orders, no matter what algorithm results are almost the same. The complexity of scheduling problems and whether they have optimization functions will result in great differences.

A large number of research data show that the distance between the optimal scheduling results obtained by rule method or heuristic method can be 30% - 150%. With the goal of minimizing the delay of orders, the APS optimized or not may always have 30 delayed orders when processing 100 orders, which is a great loss to the enterprise. Due to the limitation of the optimization algorithm technology threshold, most of the "APS" products in the Chinese market have to add a lot of manual intervention (for example, many rules are made artificially, and these rules themselves may be not good) or ignore some problems.

From the price point of view, the price of APS ranges from ten thousand or twenty thousand yuan to 12 million. The low-end products of simple algorithm for some small enterprises with simple process, from manual scheduling to APS scheduling, should be said to be an improvement, but also played a supporting role in decision-making. The actual production of many enterprises is extremely complex. APS is the highest technology content product in enterprise management software. The application of APS can improve the productivity of enterprises by several percent to tens of percent. The price of APS with truly optimized scheduling should be at least several hundred thousand or more. This shows that China's APS market and technology are not mature.

APS and MES overlap in the production function. However, the current trend is that APS and MES are integrated to achieve four closed-loop: 1. Demand forecasting and order commitment closed-loop. 2, plan and production closed loop. 3, production and implementation of closed loop. 4, order commitment and order fulfillment delivery closed loop. Form system autonomy, self feedback and self decision.

Statement: Based on online public information, this article collated and analyzed the adaptation, for reference only. Main source of high tech robot network:

Prev: [Dynamic] Electronics industry: the consumer electronics sector is coming back to life. Next: [Market] 2018 China industrial robot market performance forecast Returns List