DO WE NEED A NEW MODEL FOR SUGAR PLANT MANAGEMENT SYSTEMS?
Alberto E. Damasio
(ORSI, Italy)


1 The need for automatic help in taking decisions on plant

 Over the past few years, an increasing expansion of the "global market" and an ever-growing level of competitiveness have forced companies to critically revise both their production processes and their organisational make up. Today, a successful company is able to manufacture high-quality products with lower production costs, integrating and assimilating new production technologies into its structure to achieve and maintain a leading position in its specific sector.

In some cases these revisions lead to plant shutdown, plant merging, and plant relocating in Eastern or Southern countries. Nevertheless to manage this changes is not a trivial issue and is not always a viable solution. Therefore in some other cases business automation and process automation are selected as a way to reduce costs.

BUT to correctly evaluate costs while keeping a cost-effective product quality level, it is evidently more and more necessary a technology investment into an effective link between real-time systems (plant controllers, machinery etc..) and business management Enterprise Resource Planning (ERP) products.

Moreover it is crucial for the Company management to select a system that can support the Company through its evolution and can be flexible enough to cover all the company's needs while offering a robust and modern technical architecture to avoid becoming quickly obsolete.

Nevertheless, when the link is established, the impressive amount of data produced by the combined plant and business systems can rapidly overwhelm the decision capacity of plant managers. Such data absolutely need to be organised and presented the proper way.

Moreover the rules that must be followed in order to co-ordinate the economical strategy and the technical strategy running the company must be defined into the system and must be clearly comprehensible to everybody.

Simplification is needed. A top-down approach is needed, and each person must be given the appropriate detail in order for managers not to loose the overview of the overall context.

Therefore the structure of a company is generally reflected into a Business Model, a simplified structure identifying the internal relations and mechanisms of the enterprise. This schematic approach helps managers to keep the attention focused on key relations instead of being confused by the specific and particular complexities of the day-to-day operations.

Luckily Information Technology (IT) evolves rapidly and can greatly help in managing big amount of data and automating basic management rules. IT tools that link plant and offices are currently identified as MES (manufacturing Execution Systems) tools.

Making the right choice at the right time is one way to reduce costs, and MES helps a lot to do that.

MES is not the tool to create a completely automatic plant. This is a utopia of a few years ago that failed in being implemented in real plants.

MES is instead the tool for giving an appropriate support to decisions, executing appropriate data queries and presenting information at the appropriate time on the appropriate screen at the appropriate time, reasoning on updated real-time plant and company status.



2 The Supply Chain Model and the REPAC Model

 The theory of manufacturing management has deeply analysed the entire logistics flow from supplier to customer in order to find an ideal architecture able to react as quickly as possible to variations in customer demands.

The Supply Chain concept provides a representation of production and delivery schemes. In this model the MAKE process must be planned and considered as part of a global process, also involving Supply and Delivery. The efficient co-ordination of all the phases taking part in the Supply Chain Model allows an optimised management of the business.

AMR Research, the manufacturing industry's leading market research firm, introduced in 1992 a business model based on three "layers" of automation: Planning, Execution, Control.

Companies were used to manual procedures (paper and manual works) for filling the gap between the ERP systems and the traditional control processors. AMR highlighted the need of a new class of software products able to fill this gap: they were called MES (Manufacturing Execution System) products.

The Manufacturing Execution System Association (MESA International) has identified the following activities as those which could be included in a full MES implementation:

A good MES system must grant to be pro-active, being able to take actions automatically in response to specific plant conditions.

MESA, following the results of a study conducted on user companies, has listed a series of benefits in using an MES system, such as:

In October 1998 AMR Research refined the original MES model with a new one called REPAC which addresses all the processes required to operate the plant and co-ordinate all factory activities with the rest of logistic flow.



READY: readies production for the introduction of new Product lines or for modifications to old ones, allowing the definition of the most suitable plant configuration to manufacture the new/changed product. It also administers product and process improvements.

EXECUTE: executes an optimised production plan; it provides PROCESS with the suitable machinery configuration

PROCESS: includes all the features needed to physically produce products, including process automation and control, too.

ANALYZE: allow the analysis of meaningful data from all sources and therefore the evaluation of production performances, product quality, process capability and regulatory compliance. Data is also made available to the ERP, Supply Chain Management System, Customers and Suppliers when needed.

COORDINATE: co-ordinates plant operations with the enterprise and the supply chain. It defines the optimised sequence of plant activities to meet the production requirements defined by demand management.

This new model highlights that manufacturing is a continuous and cyclical flow of actions and information inside the company.

In 1998 Orsi introduced its CUBE-MESAD (MES ADministrator) tool, a software package to analyse, deploy and administrate MES projects according to the REPAC paradigm.



3 Usage of the Microsoft Distributed interNet Architecture (DNA) on plant

Today Personal Computers and Windows NT operating system are de-facto standards. Companies which want to be really distributed must have an efficient communication through local area and wide area network. Standard Client/Server Architecture is no more enough; when servers can be addressed by a virtually unlimited number of local and remote clients, applications must be protected from slowing down. Moreover remote clients must be light in order to run on portable computers and must be dongle-free for maximum portability, while local fat rich clients must always have a guaranteed fast approach to data.

A new control structure is emerging: the Three-Tier Architecture. This new type of architecture considers the following primary logical division among processing systems:
  • Data Servers
  • Application Servers for Business Logic execution
  • Presentation Clients


Data Servers are dedicated to collection, storage and elaboration of data. Application Servers are involved in the algorithmic part of processing and Presentation Clients are used for display and are often developed using Intranet technology such as Microsoft's ActiveX or Java.

The old "fat client" with the complete installation of software packages is no more needed.

This new DNA Technical approach can be perfectly matched with the REPAC Process approach to solve the MES problems on a Sugar Plant. At the centre of the continuous and cyclical flow of actions and information represented by the Business Processes categorised in the 5 REPAC groups, the real value of a company is the organised data resident on the Data Servers and the organised rules resident on the Application Servers, dispatching orders and information all around the enterprise.



4 Integration of process control and production management

When considering cost reduction, a correct MES solution must therefore provide a model for the plant. As a matter of fact cost reduction must not only consider direct costs (labor cost, material costs etc…) - that are normally simple to manage - but all hidden costs like: While direct costs can be reduced very effectively (relocating plants in developing countries, finding more efficient suppliers, etc…), hidden costs cannot be correctly evaluated without a model of the plant production.

What a MES system does is not (only) record costs but encapsulate the rules and the knowledge needed to reduce such costs. Once the rules are encoded in the MES package, they can be exported to different plants or extended to plants of the same companies in other countries.


Very often the current cost situation is not completely known: To record and manage all these production performance-monitoring data is normally an interesting by-product of a MES application, but is not enough.

The final goal is to help managers in defining the rules to reduce these costs.


5 Developing a MES application in sugar enterprises

In realizing a MES application, 3 phases can be identified: 5.1 Analysis Phase

Object Oriented Analysis (OOA) is a key word for the Analysis step. Human beings automatically create a model of the reality where they live. Plant operators and plant managers reason on the plant, unconsciously managing objects like: This must be effortless reflected into the MES system. This first important phase is normally done through various iterations of model implementation and model validation steps; during this phase the key factors for reducing time and money are: 5.2 Deployment Phase

In the Deployment phase, openness is needed. Very often the control environment in current sugar plants is highly heterogeneous, also because quite a lot of plant subsystems are bought from different suppliers, each one with its own automation package. Integration of the MES system with any existing DCS, single loop or other lower control devices is a must. Moreover integration is needed with upper level packages (warehouse management systems, credit/debt ERP systems, …) and third-party same-level systems (such as Laboratory LIMS packages).

5.3 Administration Phase

During administration, flexibility is needed. On-line modification of the logic (rules) connecting all the application modules building up the system architecture must be performed in a quick and simple way directly by plant responsible or shift managers.


6 The REPAC approach for sugar applications. An example: the sugar silos & packaging area

During its 17-years-long experience in automation of sugar and sugar-derivatives companies, Orsi performed a number of optimisation studies for a great number of sub-areas or different business processes that can be identified inside a sugar enterprise. Some parts where major benefits could be expected from a model-based approach to integration of process-control & production-management, include:
The last area is one of the most interesting, with some peculiarities like:
  • The sugar packaging department works all the year long, while the plant very often works in campaign
  • The sales department of the company is tightly linked to this processing area; a slow-down on sugar production does not immediately affects company business performance, while a technical problem in the packaging area is immediately reflected on the relationship with the sugar-company customers
  • Sugar packaging very often buy bulk sugar from more than one production plant and sell packaged sugar to more than one warehouse; a problem in this area can really affect all the company
  • Sales and marketing strategies (like special promotions, new products, business performance, quick delivery etc…) of the sugar company affects mainly the sugar packaging production area. The plant production is normally protected from such sales disturbance by the presence of a big buffer (the sugar silo)
  • High management of the sugar company is particularly committed in investing to solve problems in an area of high impact on the deliver part of the supply chain
Step 1:
The analysis starts with the introduction in the MESAD (MES ADministrator) tool of the current physical model of the plant. A group of objects are identified:
  • Main Silos with different sugar qualities
  • Weighing bridges
  • Intermediate tanks
  • Packaging machines
  • Palletisers
  • Mills
  • Warehouses
And material flows (through pipes, belts, lorries, etc…) are graphically introduced in the system.
Step 2:
all database or information sources present on plant or in offices are identified:
  • ERP databases (standard ERP/MRP systems - like SAP, Baan, BPCS, JDEdwards - or proprietary ones)
  • Control Systems (Scada, PLC, DCSs, of Orsi or third party providers)
  • MES databases (even if not yet present)
  • Manual paper-based files (to be probably replaced)
  • And data flows in-between this system (messages, records, reports, manual telephone call through human intervention etc…) are graphically introduced in the system.
Step 3:
In between data flows event triggers are placed, in order to automatically launch management rules when data are transferred from plant to offices or vice-versa. Each trigger is then linked to graphical rules that, when executed, orderly launch the different business processes that co-ordinate the packaging area with all the company functions.

On top of such graphic representation of the current situation a proposed layout of the future plant is overlapped through a BPR/BPI activity; benefits are identified.

For example the introduction of a label printer to insert a barcode on each pallet or on each carton box of 20 packs or on each single ½ Kg box would affect: Each single task will be implemented using any software programming language or hardware device that the sugar company may already have.

Orsi can provide pre-packaged modules already developed using its own development environments: Cube-MESAD will launch and co-ordinate such tasks. Therefore according to the REPAC task organisation: In the Execute phase typical problems that need to be solved are connected to management of the picking list received each day from the sales department (containing the data of the products brands and packages to be delivered on pallets or bulk): In the Process phase the system will acquire data from plant and will launch process tasks like: In the Analysis phase typical problems are related to the management of quality data and performance data: The Coordinate phase closes the loop with the business management of the company, trying to match production results with business needs, in order to refine and retune next production launch.

During this phase it is important: These last two activities clearly tend to automatically close the gap between the Make function and the Source and Deliver functions in the supply chain model of the plant.


7 Conclusions

The REPAC approach greatly simplifies automation control and management applications.

Systems according to these concepts have been developed with the Orsi CUBE product at various Sugar Plants all over the world, in major companies in Italy, United Kingdom, Spain, Saudi Arabia, Korea, Canada, USA, Chile, Mexico, Guatemala, etc., that already trusted the ORSI technology.
Back To Library Page Top