How to implement ERP in a machine-building holding company

— What was the business target of implementing an ERP system?

— The question was: How can we improve the efficiency of the plants and the trading company? What are the ways — comparable in results and investments — that can compete with an ERP system? We had discussions and brainstorms and analyzed the existing world practices in companies with similar production specifics. As a result, we came to the conclusion that implementing a full-featured ERP system embracing industry best practices for business processes is the right way to go. Now, at the end of such a long and difficult journey, we can see that the changes accompanying the implementation of the system are definitely having an effect.

I joined the HMS group with my colleagues in 2007. We have conducted an audit of the IT systems of all our enterprises, developed and adopted an IT strategy. One of its key points was the implementation of an ERP system. One of the largest manufacturing enterprises from the Industrial Pumps Division and a trading company were chosen for the pilot project. In the trading company, it was decided to automate the main activities from receiving an order to shipment — almost everything except CRM processes — making maximum use of the functionality of the ERP system.

The selected plant produces industrial pumping equipment manufactured in accordance with individual customer requirements. There is also a small share of small-scale production and custom-modified products. A significant share is the production of unique equipment designed to order. The enterprise that we have automated produces about a thousand units of a diverse product range per year. The processes of production support, work planning, procurement, evaluation of possible delivery dates for critical materials and components, and coordination of design and production divisions cannot be handled manually given such volume and complexity of products. Prior to the PDM and ERP introduction, the enterprise used separate systems, utilities, and tables, which partially automated the operations. Meanwhile, without uniform reference books and a unified coherent system, its efficiency is low. There are a lot of errors affecting the timing and production costs. Well, the failure to meet order production deadlines is also a whole lot of money. In general, we understood what tasks needed to be solved. The only question was that the implementation of the ERP system should be accompanied by an increase in the efficiency of all processes and not vice versa.

— One of the traditional issues of implementing ERP systems is to implement “as it is done in the system.” It means to use the best practices as proposed by the vendor or to automate the existing processes. So what did you do?

— For us, this is the most difficult project, which has entailed a lot of changes. In fact, we had two consecutive projects of implementing one system at one plant. We had enough endurance and perseverance to abandon the wrong approach to implementation that we used at first, and on the second attempt, we did everything the right way. We concluded from this experience that the introduction of an ERP system exclusively for the automation of existing production processes is a meaningless task. It is necessary to replace the old business processes — even if they are already well-known and native but have become completely ineffective — with the best practices, and only then there will be an effect… In theory, everything was clear before but we were once again convinced of it from our own experience.

When we selected Infor LN and identified the contractor, we planned to implement best practices. The management consciously resorted to such a move. This beautiful idea collided with reality a few months later. The attitude of production to implementation is always very complex: do not interfere with work, do not break processes, and do not disrupt the plan. Meanwhile, the attempts to automate the existing practices often turn into the automation of inefficient work.

Implementing best practices is a big challenge but it is precisely this that increases efficiency. Without this, it is not clear why one should implement an ERP system. It is only getting worse as people are doing everything the same as before and only adding additional information in electronic form. Without changing what they have been doing so far, there is no way they can save on the amount of work being done. This was also the case with us at the beginning — this was happening while we were just going after the production. Changing the paradigm and approach to business processes is a separate challenge in each case.

Gradually, we convinced both the management of the enterprise and the management of the business unit of what exactly needed to be changed at the plant. If we had taken the path of transformation of business processes and transition to best practices from the very beginning, we would have achieved the results much earlier. Our partner, IPL Consulting, very persistently urged us to implement the best-known practices, but at the beginning of implementation, we essentially followed the production and design-engineering divisions. As a result, a lot had to be redone.

— Please give us an example of a change in production processes.

— The unit was developed fifteen years ago. It uses materials that are no longer commercially available or unprofitable to buy. Meanwhile, these materials are listed in the design documentation. And therefore, once such a unit is ordered, it becomes clear that the materials will have to be replaced. This all is based on official memos after the launch into production. It takes quite a long time and such a process cannot be automated. As a result, in addition to numerous inconveniences and delays, there are a lot of leftovers as purchases are far from being optimal.

To avoid all this hassle, there is a well-known tool — the restrictive sheet. It is an up-to-date product range with which buyers, technologists and designers, and production can work.  If the documentation was developed a long time ago, you need a clear business process for revising this scope of documents before submitting it to production. All changes to the PDM system must be done prior to being transferred to production. It seems so easy to understand, but our technologists, designers, and production have never worked this way. Optimization of the business process is met with resistance and requires a decision at the level of the management of the business unit. The very changes in processes become the effect of the ERP system implementation.

— And implementation in the trading company: how did the project go?

— In the trading company, we immediately got a like-minded person from the business side. She set the task herself and she herself participated in the project. All the other employees, whether they wanted it or not, participated in the project exactly as it was needed. So we achieved a quick effect in the trading company. The company switched to unified product directories. All work with specifications — from discussing them with customers to the approval of the final product structure with production — was automated by means of the ERP system. The coordination of supply chain elements was also included in the ERP system loop. Over time, it even included a difficult to formalize pre-order discussion of future engineering solutions.

The trading company has been working in an ERP system for quite some time now, the processes have become more efficient than before, and the employees like to work this way. Nobody asks the question of why we did this anymore as everyone takes this tool for granted.

— How did the project go at the plant?

— The plant directors usually have down-to-earth tasks as they need to fulfill the plan, and upgrading production facilities for them is a more obvious way to improve efficiency than implementing ERP or other business applications. Only the appearance of like-minded people at the level of the plant director, the management of the business unit, and the company management helps to speed up work and reduce resistance.

Nevertheless, we had to automate some of the business processes managed by particularly recalcitrant units, as they existed at the time. Practice showed that this was not the right way, and later we had to redo everything.

The implementation of ERP at a machine-building plant becomes much more complicated if the PDM system is not fully implemented. The sequence of actions should be as follows: first PDM and then ERP – then the process makes sense. Initially, we started implementing both systems simultaneously. Pretty soon we realized what exactly we needed from PDM and in what volume, for production and in order to reduce the labor intensity, and efficiently implement PDM. We suspended the implementation of ERP and resumed it after we had sufficiently implemented PDM.

Creating documentation in itself is a goal of designers and technologists, but it does not always coincide with the interests of production. Sometimes this goal is important in itself. For example, when products are supplied to nuclear power plants then particularly stringent requirements are imposed on the documentation. However, there are many other areas where the customer does not need perfect documentation but a good enough product. In this case, we need a different approach, and we managed to find a compromise.

— What was your approach to the PDM system implementation?

— The HMS group has its own institutes, scientific and technical centers, there are design and technological units right at the enterprises. These are hundreds of specialists who create new products, modify existing ones, develop production technologies and do much more.

At the enterprise where we implemented the project, we managed to make sure that the electronic version, and not the paper version, was considered the original design documentation. The validity of printed drawings is limited. It is impossible to use them after a certain period because significant changes could have been made to the electronic primary source during this time.

The designers and technologists estimated how long it would take to prepare technological documentation for equipment that already had paper documentation. According to the requirements of the Russian Unified System for Design Documentation and Unified System for Technical Documentation, it would have taken about 80 years. Then it was decided that the documentation to be entered with the details that are necessary for production and in compliance with the customer’s requirements. In addition, to optimize the process, we described standard technological operations and implemented other optimizing approaches and procedures. Both Search, which we use, and other PDM systems have this feature. It was very difficult to convince technologists to create documentation exactly in the volume that is needed for production and, if possible, without wasting efforts on what is not in demand in production.

As a result, now all design and technological preparation at the plant is carried out through the PDM system. The plant management, designers, and technologists believe that working in PDM has become noticeably more convenient and efficient than before its introduction. It took us about two years to enter data on a large number of products. After that, the positive effect of using the system in all divisions became noticeable. Everything has become faster, easier, and more efficient than before. But it took an effort that could be described as “heroic.”

My advice to those who follow the same path is to not leave the technologists and designers unattended for a week or so as they need to be given a boost. We need drivers for this project at the enterprise. This is important for any project, but because of the huge amount of documentation the PDM implementation is not a quick process, having such a driver in place becomes a very important factor.

— How did you create a unified reference data system?

— I have already had experience in solving such a problem at a large holding company. At each enterprise, we have created a small reference data group of no more than four people. They deal mainly with the product range and reference books.  There is a management group for reference data. It has been reduced to the head who develops regulations, monitors difficult cases, and resolves some complex situations with enterprises.

We created a classifier and all the products that needed to be systematized — finished and spare parts — were codified by the factories. After several iterations, we ended up with a single common directory used by everyone, and we have been supporting it for many years already.

The reference data system is implemented in Infor LN. Additionally, we have developed a special interface – the reference data portal. All the product range is stored in a single database and it is used by enterprises. 

— Who supports and develops Infor LN?

— From the very beginning, it was clear for us that we needed our own competence center. We created a working group for implementation as we realized that over time our employees will be engaged in support. I have always had such an approach that we can implement solutions by the contractor while we need to support them on our own. It is extremely difficult to support something without taking part in the implementation. The main part of the ERP and PDM development and support is carried out by our employees. If necessary, we involve specialists from IPL Consulting, with which we cooperate almost from the start of the project. In the active implementation phase, we had four or five contractor consultants working at the same time.

— How do you form work orders?

— A work order is needed both to determine the scope of work and for subsequent calculations in the case of piecework payments. We use a bundle of Infor LN and 1C:Payroll and HR Management.

The technological route is formed in accordance with the information received from the PDM. After the Infor LN registration of execution of a group of operations of the production order created in the process of the ERP system planning, the work order can be closed in 1C:Payroll and HR Management, where the payroll is subsequently calculated based on this data. The company has a piecework system, and until employees have registered the completion of the operation in the production order of the ERP system, they will not receive their money. As a result, problems like “the work is done, but nothing is registered” have almost disappeared.

The system has the concept of “work centers” – this is a group of similar machines. In the system, we not only keep inter-workshop accounting and planning, but where necessary, we register the entry and exit of the order for each group of machines. So far, we deliberately do not use operational accounting as we do not need such details. Meanwhile, technological information on operations to the work centers is detailed in the system and we carry out accounting by groups of operations. For example, out of ten consecutive operations on one work center, a final reporting operation is allocated. Upon registration of this operation, the system automatically closes the previous nine.

— Do you conduct production planning in the ERP system?

— We currently do full-scale production order planning in the ERP system for about a six-month period. At the same time, we managed to set up various planning rules – both for parts and assembly units.

The plant planning service sets a production plan on a basis of Infor LN data, its expert knowledge of the duration of production cycles, and the actual load of critical equipment.

The next step is to automate the planning processes for active production orders with the account for the actual execution of production orders and utilization of production capacities.

To solve this problem, we need more accurate data on technological routes. Unfortunately, many standards were developed a long time ago and primarily for the purpose of calculating the piecework remuneration of employees. The parameters used in such standards do not always correctly reflect the actual production cycles. We are working on updating these technological routes and will definitely get an added value from the automation of planning processes.