An Improved Recycling Machine

29971 FUSE DEMONSTRATOR DOCUMENT Application Experiment Number 29971 An Improved Recycling Machine Microcontroller technology improves efficiency, c...
Author: August Farmer
6 downloads 4 Views 437KB Size
29971

FUSE DEMONSTRATOR DOCUMENT Application Experiment Number 29971

An Improved Recycling Machine Microcontroller technology improves efficiency, competitiveness and feature count

Blossom Design Ltd

TTN: University of Glamorgan Commercial Services (UGCS) Ltd

Page 1 of 311

29971

AE Abstract Blossom Design Ltd is a small family owned engineering company. Formed in 1965 as a sub-contract machining facility for local industry and employing only one person, it now sells its products throughout the UK and exports into other European countries and America, employing 10 people and with sales of 650 kEur p.a. Over the last 10 years the company has developed a range of specialised recycling machinery for application in the recycling of timber, plastics, paper, wires and cables, and other materials for a wide range of industrial sectors. Generally, the company’s shredders are used to convert bulky materials into a more compact form appropriate for recycling and the specialist knowledge and expertise possessed by the company enables it to tailor design the control system for each machine to suit its particular application. These recycling machines are marketed through a specialist sales office (Granutech Saturn Systems Ltd) in Bolton, UK, of which the managing director of Blossom Design has a share holding. The control applied to these shredding machines currently is simple, and consists of the use of various over limit switches and a Programmable Logic Controller to slow or reverse the speed of feeder rams and shredding rotor. The wide range of applications that these recycling machines are applied to means that the control process is not optimal, prevents the development of a universal shredding machine, requires each machine to be set up and commissioned in the customers premises by the company service engineer and often requires the company to develop new control programs for various applications. The objective of this application experiment was to apply microcontroller technology so as: • To develop an universal shredding machine based on the following technical improvements. (i)

To allow the selection of control algorithms by material type.

(ii)

To increase the number and range of sensors fitted to the recycling machine, and to select the appropriate range of sensors depending on the materials being processed.

(iii)

To apply more complex control algorithms to improve the productivity of the machine.

• To provide features for the integration of the system into up and down stream processing plant. The cost of the AE was 47 kEur and its duration was 11 months. These improvements will enable the company to improve their technological position relative to competitive machinery and allow them to hold their market share in the growth market place for environmental processing/recycling machinery. This will result in additional sales, and an expected payback period of 16 months for the application experiment costs. The return on investment over a 4 year period will be 600%, or 390% when industrialisation costs of 26 K Euro are included.

Page 2 of 311

29971

Keywords The following keywords are relevant to this application experiment: 1 2 3 4 5 6

Shredders Size reduction Recycling Microcontroller Grinding Granulation

Signature 2 0192 555 0120 2 3720 1 29 UK

Page 3 of 311

29971

1.

Company name and address

Blossom Design Ltd Blossom Works New Inn Pencader Carmarthenshire SA39 9AY Tel: 01559 384 303 Fax: 01559 384 398 Email: [email protected] Contact:

2.

David Wyatt (Managing Director).

Company size

The company currently employees 10 people in the manufacture of these machines and equipment (7 in manufacturing, 1 in design and 2 in administration) and achieves sales of over 650 kEur per annum. The company is privately owned and is not part of a larger organisation.

3.

Company business description

The main activity of Blossom Design is the design and manufacture of specialised recycling machinery for application in the recycling of timber, plastics, paper, wires and cables, and other materials for a range of industrial sectors. In addition to this, it has a range of specialised saw-milling machinery and provides a spares and servicing facility for both local industry and its existing customers. The approximate sales revenue distribution from recycling machinery, saw-milling and spares/servicing is 80:5:15 respectively. The company is privately owned and is not part of a larger organisation. Its range of recycling machinery is marketed by a specialist sales office (Granutech Saturn Systems Ltd) in Bolton, UK who sells its products directly to customers throughout Europe. 85% of the company’s sales are within the UK, 10% is achieved in Europe and the remaining 5% is exported to North America. Blossom Design manufacture recycling equipment for a range of materials. The Prodcom standard industrial classification of company activities identifies the following industrial sectors as appropriate to Blossom Design: Prodcom code 3720 (Recycling) and Prodcom code 2950 (Machinery for Special Purpose Uses).

Page 4 of 311

29971

4.

Company markets and competitive position at the start of the AE

Blossom Design has undertaken a radical change in the products that it manufactures over the last 10 years. During the second half of the 1980's, the company experienced considerable difficulties in their established market for saw-milling machinery. This was due partially to the fact that saw-milling is a mature industry and is generally a depressed market place, but more importantly it was due to a growing trend for its customer base to bypass small specialised engineering companies. Instead they purchased automated and integrated production plant from the more developed industry in the Scandinavian countries. In recognition of these circumstances, the company looked around for another product to design and develop and finally decided that size reduction machinery for the growing recycling market would best complement their range of existing skills, expertise and resources. Blossom Design still manufacture saw-milling machinery, but it only accounts for approximately 5% of annual turnover. The recycling machine marketplace is complex in terms of its structure, but it is a strong growth industry and environmental legislation continues to maintain pressure on all industries to invest in recycling technologies. The complexity therefore is derived from the characteristics of the wide range of materials requiring reprocessing, which has resulted in specialised machines with defined characteristics for each specialised application. The recycling industry has experienced strong growth over the last decade as new technologies have been developed to enable an increasing variety of materials to be recycled economically. More recent European directives on recycling of packaging, domestic, automotive and other waste streams has resulted in yet further pressure for industry to recycle rather than discard. Large "Blue Chip" companies have led the way by gaining ISO14001 environmental accreditation and the technologies and procedures thereby developed have been used to enable other smaller organisations to improve their housekeeping and recycling procedures. With regards to product life cycle analysis, the recycling industry has reached the strong growth phase and therefore there is increasing pressure for product development to keep pace with the evolving market requirements. Blossom Design sell machines to both the end user and to other environmental engineering companies which require a shredder as part of their integrated recycling plant and they have manufactured machines for each of the following reprocessing machine market sectors: • • • • • • •

Wood (hardwood, softwood, veneers, particleboard) Plastics (hard and soft, lump & film) Chemical (solids & liquids) Medical (soft tissues, clothing, instrumentation) Soft Metals - copper and Aluminium Glass (solid & in fibre form) Demolition Waste.

The majority of the company’s Shredding Machines are generically similar, but currently require extensive controller adjustments to accommodate each material. The company’s UK market share is approximately 5% of the total recycling machine marketplace of 13 MEur. The competing companies include: IRS. Vecoplan. Page 5 of 311

29971

Weima. Lindner. Untha The largest supplier in the UK is Weima who manufacture a large range of recycling machines for the industry and can provide fully automatic and integrated systems. They have approximately 40% market share in the UK; the remaining market share is distributed approximately equally amongst the other competitors. This is illustrated in Figure 1.

5

13

Blossom Design

10

15

17

IRS. Vecoplan. Weima. Lindner. Untha

40

Figure 1: Market Share Levels All of the competing products in the company’s marketplaces are limited to the recycling of a small range of materials. At the start of the AE, none of the competing products incorporated any instrumentation in their products, and all of these products used simple PLC type controllers. These technical features limit the ability of the products to reprocess a range of different materials, and require manual supervision. However, at a recent recycling exhibition in Cologne, one competitor showed a video illustrating one of their machines with a microcontroller unit! The price of the company’s recycling machines varies with the physical size of the unit. The price varies from 15 kEur for the smallest machine to 100 kEur for the largest capacity product. The installation costs for the product can result in a significant additional cost for the customer, and in many instances the cost for infeed and outfeed conveyors and their controls can often exceed 120 kEur. The total contract value to Blossom Design can therefore, in some cases, be double the basic price of the equipment. Sales over the 3 years prior to the AE were as shown in Figure 2. Despite substantial sales discounting taking place in 1999 due to increasing competition from abroad, relative profits were maintained, as shown, but this was against a downward trend in both sales and profits.

Page 6 of 311

29971

100 90 80 Sales

70 60

Profitability

50 40 30 20 10 0 1997

1998

1999

Figure 2: Relative Sales and Profitability in recent years (1997 shown as 100) Competition in the rapidly developing recycling market is based on technical advantage. The machines are sold into a technical market, and therefore the provision of improved technical performance is valued by the end customers. Evidence that competitors are rapidly developing their product, including a microcontroller, has been provided from attending a recent recycling exhibition in Cologne. These technological advances must be matched if the company is to avoid being squeezed out of the market place. The technical advantages that the application experiment will provide Blossom Design are significant, not only because the improved product will provide a more efficient and productive operation, but also because it will enhance the perceived value of the product to the customer. Moreover, it will eliminate the need for service engineers to commission each unit and will also enable the design of the company’s control systems to be more standardised, which will further reduce manufacturing costs. The provision of an RS232 connection will further enable remote diagnostics to take place to enhance after sales service capability and provide confidence to customers that Blossom can rapidly respond to their call for advice or assistance.

5.

Product to be improved and reasons to innovate

Blossom Design’s range of DragonT M shredders are designed to reduce solid or bulky waste residues into smaller granulated particles suitable for further recycling. Each of the products consist of 3 basic functional components. These are: 1. A hopper into which the waste material is dumped. 2. A hydraulic ram which pushes the material into the cutter. 3. A rotating multi-blade cutter and perforated screen. This unit fragments the bulky materials deposited into the hopper, and particles smaller than the graticule size in the perforated screen emerge through the screen into an output container.

Page 7 of 311

29971

Figure 3: Existing Products The Dragon shredders are constructed from heavy gauge welded steel panels and frames, and the major moving components including the cutting rotor are manufactured from solid steel. The company purchases the motor, gearbox and the hydraulic control system for the machines from established manufacturers. The rotor and cutting tools are made from high grade steel, and the cutters are hardened to maximise the life of the tool before replacement. The Dragon Shredder is provided in a range of capacities from the smallest machine incorporating a 170mm diameter by 50mm long rotor up to the largest size having a 500mm diameter by 1750mm long rotor. The basic models of Dragon Shredders are D50, D80, D110, and D160 and for each model the client can specify rotor diameter, number of cutters, hopper length and motor power. The technical specification for these models is summarised in Figure 4 below. More information is available from the company’s marketing office.

Machine Model Hopper Opening (mm) Rotor Diameter (mm - max) Main Rotor Power (kW-max.)

D50 500x750 170-270 7.5-11

D80 850x1100 170-370 11-30

Figure 4: Machine Model Range

Page 8 of 311

D110 1150x1200 270-370 22-37

D160 1750x2000 370-500 45-90

29971

Fast / Slow Control

Hydraulic Ram Hydraulic Motor

Reverse Stop

Rotor Motor

ROTOR

Reverse Rotor Current Sensing / Trip Setting

Figure 5a. Block Diagram of Existing System

Figure 5b: Existing PLC-based control box

The Dragon Shredding machine incorporates a simple PLC unit which reacts to 4 relay trip inputs generated from preset levels of current sensing to ensure that the machine’s operation is crudely controlled. The control outputs allow steps in motor current for rotor load control Page 9 of 311

29971

and the reversing of the hydraulic ram only. However, the control process does not provide any analogue control of the output levels for the motor drive or for the hydraulics, and does not incorporate any sensing instrumentation to allow more sophisticated control of the shredding machine. This limits the operational performance and range of commercial application areas for any one setting of the controls for a specific machine. For example, in the plastics industry, the shredder control system has to be set differently for hard and soft plastics and different again for low melting point plastics in order to achieve maximum productivity. The objective of the application experiment was to remove this technological limitation and to produce an universal machine which would help maintain the company's market share in this technology sensitive industry which spans most industrial sectors which produces a waste stream. A list of parameter improvements is shown in figure 6, comparing the features available on the existing and new shredder machines A further objective was to ensure that this step improvement in technology enabled further developments in the future and a possibility of retrofitting at least some of these by a simple up-grade.

6.

Product or process improvements

The improved product includes additional control features to provide the following product enhancements for a range of new generation shredding machines: • • • • •

Automatically adjustable shredding characteristics to provide an universal capability. Improved productivity by automatic adjustment of the hydraulic ram and cutting rotor speed. Infeed level sensing and adjustment of the shredding characteristics to maintain a controlled flow of materials. The provision of diagnostic reports on the operator screen display if operating conditions exceed preset values or if any part of the machine develops a fault. The facilities for algorithm improvement and development, with upgrading of the new generation machines in the field.

The improved shredding machine incorporates the following sensors: 1. 2. 3. 4.

Cutting rotor motor drive current. Cutting rotor temperature. Hydraulic ram pressure. Hydraulic ram position. This is determined by a simple algorithm from ram pressure and the direction and duration of travel. 5. Infeed hopper material level. This is gauged approximately from the response of an ultrasonic level sensor mounted on the hopper. The improved product controls the rate of movement of the hydraulic ram so as to maximise the throughput of materials. This is achieved by the electronic control of an electrically controllable hydraulic valve with reversing capability. Provision is made for the future incorporation of a more sophisticated infinitely variable hydraulic speed control to further improve production efficiency. Provision is also made for the future incorporation of a two speed hydraulic rotor motor drive, which would thereby be able to double the drive torque Page 10 of 311

29971

available to the motor when processing extremely difficult materials. The following table gives a comparison of features on the existing and the new machines. Figure 6 gives a comparison of the features of the existing and improved products. Features

Existing Product

Improved Product

Control Operator Controlled Programming Features Control Algorithm Ability to adjust to accept a variety of materials Hopper level detection Jamming detection Jam clearing

Manual None -Fixed

Semi-automatic Selectable – by menu selections Complex Automatic (within limits) Available –ultrasonic sensor Automatic Automatic reversing algorithm Available LCD display Available at display RS232 link Up-grade by EPROM replacement Good Extensive

On / Off None None Manual Operator controlled

Temperature sensing Operator display Diagnostics Communications Up-grade capability

None None None None Machine re-build

User friendliness Provision for future development

Poor None

Figure 6: Comparison of features

Page 11 of 311

29971

The functions performed by the microcontroller in the improved Dragon Shredder equipment are described with reference to the functional block diagram in Figure 7. Rotor Temperature

Microcontroller

Hopper Level Sensor Rotor Current Sensor Rotor Speed

Hydraulic Pressure Sensor

Rotor Motor Control

Hydraulic Pump Control

Hydraulic RAM Position Sensor

Operator Key panel + Display

Figure 7

Serial Interface

Functional Block Diagram of the Microcontroller System

The functions performed by the microcontroller include: 1. Interfacing to a range of sensors for the cutting rotor and the hydraulic ram. The microcontroller samples the rotor current repeatedly to determine whether an overload condition could arise. The rotor speed is monitored by a shaft encoder coupled to the rotor. 2. The rotor’s temperature can be measured using a thermocouple. The temperature sensor’s input is used to ensure that the cutting efficiency is optimal, and that for certain plastics the temperature of the rotor does not rise to the melt temperature for these materials. 3. The hydraulic ram end position (determined by way of an algorithm of time and distance Page 12 of 311

29971

moved by the ram) and electrical sensors to monitor the hydraulic motor load. 4. These parameters are applied to the control algorithm to determine the optimal controls for the recycling process. 5. Infeed level transducer measurement, and the application of the information in the control algorithms to ensure the required level of machine productivity. 6. Control algorithm calculations. The microcontroller determines the optimal drive output settings for the equipment using rule sets and interpolation tables determined for the category of material used. 7. Rotor motor speed selection. 8. Hydraulic ram control by the analogue control (0-1V output) of rotor load (with provision for future incorporation of an infinitely variable hydraulic speed control facility). 9. Operator interfacing, including the provision of a LCD (Liquid Crystal Display) and an input keypad. The microcontroller performs all of the key interpretation and display generation requirements for the improved control unit. 10. The provision of a serial RS232 compatible communications facility to provide the capability for interfacing to other controller units in the final reprocessing plant, and for the transfer of statistical data to a remote PC if required.

Figure 8: Improved control panel using a microcontroller

Page 13 of 311

29971

Figure 9: Photograph of the Circuit Board Shredding technology has been available for 10-12 years and continues to evolve as the range of applications for this type of machine increases. The basic engineering design of the machine has been refined to a point where it no longer provides a limitation to application. Rather, it is the control system that has become the restriction and therefore the main focus of developmental interest. Blossom Design has evidence that the competition are now developing a similar microcontroller based control system and without this application experiment the company would be at a significant technological disadvantage that could have resulted in severe economic consequences. The ability to extend the capability of the control system and to upgrade it in the field was a major consideration in justifying the selection of a microcontroller.

7.

Technology, tools and methodology choice rationale

The technologies considered for the control of the improved Dragon Shredder recycling machine included microcontroller devices, PLC’s and single circuit board PC’s. Alternative microelectronic device technologies were found not to be appropriate for this application experiment for the following reasons:Discrete device technologies did not offer the degree of flexibility required to implement the control algorithms and would necessitate a large PCB to accommodate the high number of devices required. -ASIC and MCM bare die based solutions were not justified because there was no need for ultra small size, and the non recurring engineering (NRE) costs were not justifiable given the low annual volume of shredding machines produced.

Page 14 of 311

29971

-FPGA implementations of digital state machine controllers is an inflexible and time consuming method of realising the required control algorithms, and is more costly in terms of development time and device costs than the selected microcontroller option. PLCs were currently used in controlling the existing range of Dragon shredders. The strength of PLCs is their ability to control simple sequence of operations as are implemented in the existing product. The improved range of equipment requires more sophisticated control methods. The only option available in the PLC to accommodate the more advanced control requirements was the use of PID control modules. In this application PID controllers did not meet the system requirements because: 1. PID controllers operate correctly when the process to be controlled is stable and predictable. The shredding machine is subjected to a range of different size and grade materials, and materials with different shredding properties. This results in the presence of non-linearity’s and time varying process parameters, which prevent the correct operation of PID controllers. 2. PID controllers operate on a single process parameter, and control this single parameter. In the case of the improved Dragon shredder machine, several parameters must be monitored and controlled together. PID controllers cannot achieve this interdependency, and therefore a higher level of operator intervention (‘supervisory’ control) is required when PID controllers are used in this situation. 3. A universal shredder machine places different levels of emphasis on monitored parameters for each application. The provision of PLCs with a PID controller for each variable introduces a higher level of cost than is necessary and still has a less satisfactory route for future feature enhancement. PLCs therefore did not offer the degree of flexibility and cost effectiveness possible with microcontroller device technology. The use of PC based cards require several plug-in peripheral modules to provide the required functionality and additional circuit board development for the interface to certain sensors. The reliability of the resulting assembly consisting of several interconnected circuit boards and the PC cards was considered unacceptable in the severe environment encountered in terms of EMC and vibration. The application of fuzzy logic control to the control of the recycling machine was investigated with the subcontractor and TTN. A review conducted with PLC suppliers at that beginning of the AE indicated that there were no fuzzy logic tools available for algorithm development using PLCs. There were tools available for microcontrollers, and this biased the approach towards the selection of a microcontroller. However, as the company did not have the intuitive rules to develop these algorithms, and because the technology step from nonmicroelectronic capability to fuzzy logic system development was considered too high for one design project, fuzzy logic developments were postponed at this time. In contrast, the design of a dedicated microcontroller circuit allowed the inclusion of EMC design protective measures, and enabled the control of a multiple variable input system. The application of microcontroller technology had several advantages in contrast to PLC systems, and allowed more complex control algorithm implementation. These factors led to the selection of microcontroller device technology as the preferred implementation. Several possible microcontroller solutions were investigated. The three devices identified for further consideration were: Page 15 of 311

29971

• • •

8XC552 Philips device is an 80 pin QFP package Motorola 68HC11E9 device is a 64 pin QFP. PIC microcontrollers.

In order to choose a preferred device, the three alternatives were compared against the following list of additional requirements:• • • • •

Cost of microcontroller, necessary support components and design and development tools. Compliance with current and expected technical regulations. Environmental ruggedness, physical size and packaging (with consideration of small scale PCB manufacture). Ease of field testing and servicing, availability of spares and expected product line life span. Upgrade route for future development.

By carrying out an early specification review and thus delaying the formal training session, we were able to make the final choice of component after attending the training course. During this training session, we were able to discuss the pro's and con's of the three alternatives with the training staff, who were very helpful and knowledgeable about the availability and suitability of development tools for the different microcontrollers. The choice of software tools needed to reflect the requirement to provide an integrated development environment that included: • • • • •

"C" high level and assembler programming. The ability to test code by simulation of the controller so that early development was independent of the concurrent hardware development. Emulation - so that the bench and PCB prototypes could be tested. Final programming of the microcontroller chips. Cost.

As a result of the above selection procedure, the following key components were selected: • • • • • •

Microchip PIC 16C77 microcontroller. MPLAB windows IDE simulator & editor. CCS PCW C compiler MPASM assembler ICEPIC in-circuit emulator PICSTART Plus microcontroller programmer.

Software development has been conducted using the C high level programming language. The response time of the system did not justify assembly language programming. Programme development utilised a top down approach using methodologies defined by the subcontractor. Software modules were separately tested using the simulator facilities provided by the compiler, before using the in-circuit emulator to demonstrate the correct functionality of the software in the embedded environment. The final implementation of the software required one design iteration to enable the operational code to be fitted into the PIC device. This was successfully achieved, and the use of an external EEPROM device has removed memory limitations in the final implementation. Page 16 of 311

29971

It was decided to build a bench prototype on which it would be possible to test both the hardware and software design. By using this approach, it was possible to test individual elements of the hardware design, analogue conditioning, digital input/output multiplexing, I2 C EEProm, RS232 interface, Keypad and LCD circuits, etc. By using the in-circuit emulator, it was also possible to test individual functional parts of the software, LCD driver, keypad scanning, multiplexing I/O driver, etc. This approach led to less iterations of the PCB design and manufacturing process than originally expected. When the hardware circuit was finalised, it was possible to continue software development independently of the production and assembly of the PCB. The controller’s circuit board was designed to use pin-in-hole devices wherever possible. The circuit board design included design for EMC compliance components, including power supply and input/output filtering. The performance of the prototype PCB was so good that it was decided to integrate it straight away into a standard electrical control panel and fit it to one of our standard stock machines for final functional and acceptance testing, which was thereby carried out on a real machine in an actual working environment. The design methodology applied during the application experiment were devised in conjunction with the selected design subcontractor.

8.

Expertise and experience in microelectronics of the company (prior to the application experiment)

Blossom Design’s technical capability lies in the area of recycling machine design, and specifically in the mechanical design of these products by its Managing Director. The company possesses basic machining, mechanical fabrication and assembly skills. The company had not undertaken any electronic design tasks before now and the control system design for the existing range of materials was developed using an external subcontractor. However, the company does now undertake the installation and commissioning of these controls into its machines using its technician staff. In some cases the company undertakes the electrical installation of associated plant, such as conveyor systems, at the customer’s site. Blossom Design had not undertaken any electronics system design, had no electronic design CAD tools, and had not undertaken any electronic assembly tasks. Blossom Design was therefore a first user in the area of microelectronics technology. However one of these technicians is the company employs one self-trained, with some exposure to electronic systems as a result of a personal interest and previous activities conducted by that individual on military systems at a previous company. This basic electronic servicing and maintenance experience has been developed further in this application experiment. The Managing Director and this Technician have jointly worked on this application experiment, but whilst the MD has become familiar with the operation of the prototype product, the technician has gained the most detailed knowledge of the microcontroller functionality. The MD is a Chartered Mechanical engineer, but with no prior electronic experience. The technician had no electronics qualifications nor experience beyond that of installing and maintaining bought-in PLC units. Page 17 of 311

29971

9.

Workplan and rationale

The application experiment did not run exactly in accordance with the workplan originally drawn up. Deviations occurred as microcontroller knowledge was transferred to the company and it became apparent that on the one hand there was a very large selection of microcontrollers to choose from, and on the other that microcontroller technology was much more useful than had been originally perceived. It was concluded that the choice of microcontroller was critical for the future potential of this product to be fulfilled and advice was taken from the TTN, the training consultant and the subcontractor to ensure the most appropriate choice was made. The work packages undertaken during the application experiment are described below. The resource allocation to these activities is tabled in figure 10 and the time sequencing is illustrated in the two Gantt charts in figure 11.

Activity

Company Planned Time (man days)

Company Actual Time (man days)

Subcontractor Planned Cost ( kEur )

Subcontractor Actual Cost ( kEur )

Training

8

10

1.4

0.8

Specification

20

20

0.8

0.8

Hardware Design

32

35

4.6

2.7

Software Design

45

59

6.8

8.2

Technical Mangmt

25

22

-

-

Evaluation

38

41

4.3

6.5

Figure 10: Resource deployment table Training: Originally, the formal microcontroller training course was to be held at the beginning of the project. Holiday commitments resulted in an initial delay in starting the project and with hindsight, July was not the best time for the start. However, it soon became apparent that the choice of programming software development tools was dependent on the microcontroller choice and, in turn this was dependent on the agreed system specification. In order for the formal training to be directly relevant to the application experiment, it was agreed that the system specification should be defined before the formal training was provided. This proved very successful, because the trainer provided valuable advice regarding the final choice of microcontroller (based on our system requirements) and the software development tools that would be best suited for that choice. In addition, he provided further relevant guidance that enabled additional software development to take place (using the breadboard technology with the emulator) that would otherwise not have been possible. The breadboard circuit allowed Page 18 of 311

29971

the software to be corrected as it was being developed and also enabled changes to be made to the hardware design as the requirements of the end product were upgraded. Further training was provided by the subcontractor to enable the company engineer to understand how the programme was structured and thereby how to make simple adjustments as and when necessary to such parameters as timers, current control limits, etc. This further familiarisation training was on-going throughout the project, but since the design subcontractor did not invoice specifically for on-the-job training, this was allocated to the appropriate activity and not to training. The training costs therefore reflect only the formal training course, but the duration is shown as extended. Specification: The provision of a system specification was the first task undertaken on this application experiment. The original specification of the microcontroller function was drawn up generally along the lines of the existing PLC programme. Whilst doing this, it became apparent that some features of the original programme were unnecessary and that a significant improvement in machine control was actually achieved through the simple process of reviewing the existing programme. During the 11 months period of the project, the company continued its policy of product and market development. It identified additional control features that would probably be required in the future (such as choice of language, diagnostic counters to establish the frequency of certain trip conditions, an algorithm to recognise supply voltage fluctuations and to take appropriate action, etc). The microcontroller programme was improved so that these features could be incorporated at a later date. The duration of the specification phase was therefore extended, but the resource expended did not exceed that which was planned, because activities which were not included in the original proposal were not booked against the AE. These other activities included measurements to better understand the characteristics of the existing machines and the development of algorithms and making design provision for features which were beyond those proposed for the AE. Hardware Design: The design of the microcontroller circuitry and circuit board design was undertaken by the subcontractor, with only minor collaboration by the company. The continued use of the breadboard circuit during software development reduced the time taken by the subcontractor, because of the ease of making changes during de-bugging, but this doubled the duration of the hardware design. Considerable time was taken when the completed circuit board was hooked up to an actual shredding machine in order to arrive at the appropriate values of components to provide adequate noise protection and correct analogue signal amplifications. These problems were not identified whilst working on the bench-test rig. Software Design: The prolonged use of breadboard circuitry enabled the software development phase to overlap the hardware design and thereby allowed maximum refinement of the prototype product. The software design task, however, proved more time-consuming than had been expected. Initial testing of software on the breadboard circuit was conducted on a bench-test rig. This Page 19 of 311

29971

allowed software correction to begin at a relatively early stage and as the programme was being written. Considerable time was spent in developing software for the operator interface. This accounted for a large part of the resource over-run incurred on the software. Part of this was an underestimate of the resource required for this task, but part of it was the realisation as to just how important it was to make best use of this interface. The multiplexing of the inputs and outputs required a large portion of memory and was found to be very complicated. At an early stage, it was agreed that the main programming language would be "C" in order to be able to develop the software more easily and it was also agreed to use a separate chip for data storage as it was anticipated that memory could be a limitation. This was a fortunate decision because the final microcontroller had less than 10% free memory, although once the prototype unit had been proven to work, this could be remedied. Either, many of the repetitive parts of the programme could be re-written in assembler language to free up available memory, or else memory could be re-allocated to space available in the off-chip memory. In practice, a mixture of both approaches was adopted, which solved the problem but was not an optimum solution. Despite a 30% over-run by the company on software hours and a 20% over-run by the subcontractor, the task was substantially finished within the planned duration, but with minor changes continuing throughout the evaluation phase. Evaluation: The evaluation of the completed prototype microcontroller was undertaken using one of the company's stock shredding machines. The control settings for the particular shredder were initially programmed via the operator interface and then each function of the microcontroller software was evaluated in turn. Considerable time was taken to evaluate the algorithms controlling the response of the machine to variations in rotor loading (rotor amps) and to ensure that the machine reacted in a predictable and controlled manner. Some features of the microcontroller programme, such as temperature monitoring and ram speed variation could not be tested on this particular machine. This was because it was not fitted with these options or because the variations of actual usage could not be replicated in the workshop. Therefore the inputs were simulated by means of switches and the required outputs confirmed by lights. The final testing of these features is deferred to the industrialisation process, where a machine with the complete set of options will be evaluated in an industrial scenario. This testing does not require an additional prototype control system to be developed, as the existing prototype has been designed to allow the unit to the company’s range of machines. Technical Management: The technical management of the project was seen to be crucial to its successful completion. Consequently, additional support was provided for the subcontractor (from the TTN & training subcontractor) to ensure the correct choice of microcontroller and other technical matters. The importance of an appropriate system specification was also recognised and additional resource applied to this area early on. More resource was allocated to the practical testing and evaluation phase of the project due to the company's conclusion that this was the most essential part of the AE for them. The knowledge transfer process was undertaken as follows: The basic electronic knowledge possessed by the company’s technician was developed Page 20 of 311

29971

in the first instance by attendance at a course on microcontroller systems design and microcontroller programming in C operated by the training provider, CEPE. Formal training on the selected device, programming the device using a C compiler, and the use of low cost development tools was provided. This training course provided the essential knowledge subsequently developed by the hands-on training provided by the subcontractor during the application experiment. This process conveyed a broad but basic level of understanding. The code was, however, written by the subcontractor and at least initially it is planned to make subsequent software changes through the subcontractor. Hardware development skills were developed by the day to day involvement of the technician in the design process, supported by the design subcontractor. Knowledge in device selection, circuit schematic development, and the simple use of CAD tools was gained. This interaction also assisted in the ability to apply appropriate test methods, tools and equipment. These test skills are highly important in providing the level of in-service maintenance support required or the improved product.

Page 21 of 311

29971

Task Description

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

Training System Specification Hardware Design PCB Design Software Design Test Rig Design Evaluation Technical Management PLANNED TIMESCALE

Task Description

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

Training System Specification Hardware Design PCB Design Software Design Test Rig Design Evaluation Technical Management

ACTUAL TIMESCALE

Figure 11: Planned and Actual Timescales

Page 22 of 311

29971

The company’s managerial expertise in the area of microelectronics was developed by the Managing Director’s involvement in the programme. This aspect of knowledge development was also assisted by the strict TTN reporting and monitoring regime, and by the provision of advice from the design subcontractor. The roles and responsibilities were defined as follows: The MD undertook the technical management throughout the AE. He also led the Specification phase and closely directed the Evaluation phase. As a result, the AE was well directed throughout and the customers’ requirements were thoroughly represented, both in defining what was to be incorporated into the improved product and in evaluating that it had been achieved. The technician worked closely with the subcontractor during the design phase, both hardware and software and gained a good working knowledge of the system. The continued use of a breadboard circuit enabled the hardware and software to be designed in parallel, which also helped in the familiarisation and on-the-job training process. He also built the test rig, which enabled system simulation and software development on the bench prior to full evaluation on a shredding machine.

10.

Subcontractor information

The company is situated in a remote part of rural Wales and the choice of subcontractor was limited. It was considered essential that the company chose a subcontractor that they were comfortable with and that could provide close supervision during the application experiment. Venefica Systems Ltd was chosen to be its preferred design subcontractor during this application experiment for the following reasons: • • •

They had previous experience in the area of microcontroller design and had worked with the company on several occasions on the design of the control panel for the existing shredder.. They were therefore also familiar with the product. They had experience of using microcontrollers in severe operating environments and this knowledge was considered valuable in reducing the risks of implementation. They were geographically located close to Blossom Design, which is situated in a remote, rural area of Wales and their ability to provide a reasonably prompt response to problems arising during the application experiment was seen to be crucial.

Close contact has been maintained with the subcontractor for the duration of the application experiment and the company is very pleased with the work provided. The selected training provider for the application experiment was a local University supplier. The subcontractor arrangements were satisfactory in all respects. The company entered into a formal subcontract with the subcontractor, and identified staged payments for the delivery of key items of work. The intellectual property rights and the ownership of all associated drawings and files reside with the company as part of this subcontract.

Page 23 of 311

29971

11.

Barriers perceived by the company in the first use of the AE technology

Blossom Design faced several knowledge, perceptual and management barriers to adopting microcontroller technology, including: 1. A lack of knowledge in microelectronic technology, and no understanding of the capabilities and application areas of various forms of electronic components. This formed a major barrier in terms of understanding the commitments and risks faced by the company in adopting microelectronic technology for the first time. 2. Perceptual barriers associated with the view that microelectronics technology involved high levels of investment in its adoption, requiring expensive capital investment in specialised facilities and technical expertise. This perception had biased the company away from adopting microelectronic technology to a “more acceptable” electromechanical technology, such as relays, limit switches etc. 3. Concerns about the reliability of microelectronic technology in the environment faced by the company’s recycling equipment was influenced by the company’s view of microelectronics technology as a technology suited best to clean, office type environments. 4. Lack of knowledge in terms of calculating design, development, and production costs prevented management from considering the financial benefits of the adoption of the technology. 5. Perceptions of high risk led to the adoption of alternative, albeit technically inadequate, solutions being adopted and with which they had become accustomed. 6. A lack of management knowledge in terms of specifying and controlling project implementations based on microelectronics technology increased management’s reluctance to consider the technology.

12.

Steps taken to overcome the barriers and arrive at an improved product

The steps undertaken to overcome the barriers stated above were as follows: 1

An initial appreciation of the capability and of microcontrollers was provided during the initial task of drawing up the FUSE proposal and the final system specification. The subcontractor advised what additional features could be provided by the microcontroller design and which were not present in the existing PLC programme. Subsequent attendance at a formal microcontroller training course provided further information about microcontroller applications. As a result of these actions, the company is more knowledgeable about the capability and applications for microcontrollers and is actively seeking new products that may be upgraded by the incorporation of this new technology.

2

The investment costs in terms of acquiring development tools and the expertise was quantified during the FUSE proposal stage, and had corrected the company’s incorrect perceptions of start up cost problems. Blossom Design has purchased the necessary

Page 24 of 311

29971

development tools required for designing microcontroller systems and has shown that a microcontroller based shredder control system incorporating significant improvements over previous technology does not cost significantly more. The application experiment has proved that microcontroller based control systems are affordable for small manufacturing companies and Blossom Design are no longer restricted to electro-mechanical control devices. 3

The subcontractor provided examples of existing microcontrollers operating in severe environments to provide the company with the confidence it required in the reliability of this technology. The new prototype control system has so far proved absolutely reliable and will in time reinforce the initial confidence provided.

4

The development costs were quantified at the FUSE proposal development stage, and the company therefore was able to quantify the design and development costs at the start of the application experiment. The device costs were also estimated and provided reassurance that the manufacturing costs would not exceed the current implementation. The application experiment has allowed Blossom Design to carry out a microcontroller development programme and successfully build a prototype control system at a cost approximately equal to their standard product. The company has proven that the production costs of this new technology are affordable.

5

The quantification of the financial costs reduced the perception of high risk associated with this development. These factors, and the impact of emerging competition using microcontroller technology would have encouraged the company to reconsider this risk and undertake the development independently. Whilst the availability of funding under the FUSE initiative has enabled Blossom Design to accelerate this development process the risk perception had been lowered to a level commensurate with project funding consideration.

6

The support mechanisms from the TTN and the subcontractor during the application experiment has provided valuable experience for the company’s management team. This reduced the technical management barrier, and the company no longer regard microcontroller technology as outside their grasp.

13.

Knowledge and experience acquired

Blossom Design have developed a broad range of new capabilities, including: 1. System requirements specification and subcontract preparation skills. 2. Microelectronics project management skills, including knowledge on the methods of planning microelectronics design programmes. 3. Enhanced commercial skills, including the capability to financially appraise the benefits of future proposals for microelectronics process enhancements. 4. Electronics design skills in the area of microcontroller interface circuit design to allow the interfacing of the device to sensors, displays and actuators. 5. Microelectronic, and specifically microcontroller testing skills. This includes the use of

Page 25 of 311

29971

electronic test instrumentation, microcontroller emulator facilities, and various methods for isolating and correcting software faults in the microcontroller’s embedded program. 6. Knowledge in defining the test methods for use in the maintenance of the improved product. The primary objective of the application experiment was to provide a level of expertise and confidence of microcontroller technology from which Blossom Design might develop further applications and eventually embrace this new technology completely. This objective has been fully achieved. It was not the company's objective to become an expert in microcontroller design in one project step, or to become independently proficient in the use of a 'C' compiler for developing software coding. Rather, it was considered essential that the company be able to apply the microcontroller technology to improve existing products and to supplement its ongoing policy of product development and innovation. This objective has also been achieved.

14.

Lessons learned

The following lessons have been learnt during the period of the application experiment: 1

There are a vast range of microcontrollers to choose from and specialised knowledge is required to determine the most appropriate selection for any particular application. On looking at the various microcontrollers on the market, it was apparent that the required software development tools (and subsequent training in their particular use) could not be specified until the microcontroller had been chosen. In turn, this required the system specification to be established. It is therefore advised that replicating companies provide a detailed technical requirement specification to an independent agency (for example, a local technology transfer support agency) to confirm the optimum device selection.

2

The procedure required for developing the software design flowchart made the company look very closely at their existing control system. This resulted in the identification of a number of shortcomings that have been designed out of the new system. At the same time, new ideas for improvement were generated by the reviewing process that has further enhanced the new control system. Replicating companies should be prepared to reconsider the ‘way things are done’ in their systems when a flexible technology such as microcontrollers are used.

4

The formal microcontroller training course was easier to follow than expected. The practical training sessions were particularly useful in providing confidence in the use of the software development tools. Such training is recommended.

5

The use of a breadboard circuit design and emulator enabled software development to progress sequentially. This design approach allowed for corrections to be made easily, resulting in the final prototype microcontroller to be more fully developed than would otherwise have been the case if it had been ordered earlier. The use of breadboard prototypes and emulators to fault find and fully test code is recommended.

6

It is always necessary to try the finished product out in a real application, as this Page 26 of 311

29971

exposes the circuit board to environmental noise that would otherwise not be detected during bench testing conditions. The test process needs to include this thorough machine level test. 7

The software coding, correction and final evaluation all took significantly longer than anticipated. It is advised that this additional time is provided as a contingency in any development programme.

15.

Resulting product, its industrialisation and internal replication

The prototype controller has been demonstrated successfully, and is fully operational. Already a number of its advantages have been demonstrated in the workshop: it can detect a jam and take appropriate rather than just a pre-scribed response, it can adapt to being part or fully laden, it can adjust to harder or softer material, to larger or smaller components to be shredded and it can adjust the speed and travel of the ram accordingly. These improvements have still to be quantified in a live commercial application, but have been achieved qualitatively. The improved shredder control system is now to be extensively trialed in an actual machine as soon as the company receives their next order from a local company (in the South Wales area). Further development will be conducted to achieve the full potential of the new control system and it is therefore essential that it is installed locally so that its operation can be adequately monitored and controlled. It is anticipated that it will take another 6 months before the prototype machine has been satisfactorily trialed and developed to a stage where it can be exploited commercially. This is due to a need to ensure reliability and also to establish data to support the higher performance of the improved control system. The production model will have to be improved aesthetically and further algorithms developed for a variety of materials not yet trialed. The company has identified several recycling products aimed at niche markets within the target industrial sector for the further application of microcontroller technology. These include the development of: • • • •

Infeed and outfeed material processing equipment. Advanced material screening systems for the recycling process. Separation systems for material reclamation. Granulators for the plastics recycling industry.

Each of these product developments require the application of sensing instrumentation and advanced control systems to achieve the desired level of performance identified for the target niche markets. The product diversification potential of the microelectronics technology is potentially very large. The adoption of in house electronic design knowledge is crucial to implementing this expansion. The remaining industrialisation tasks are as follows:

Page 27 of 311

29971

(The apparent three month delay from the end of the experiment before starting these activities is due to the need to locate a suitable local customer for the initial field trials.) Task

Description

Month

1

Trialing the prototype machine at a customer’s premises in the 3-9 vicinity of the company

2

Collecting data on its performance efficiency

3-5

3

Further improving the control algorithms if necessary

5-9

4

Establishing confidence in its reliability

3-9

5

Refining the operator interfaces, using an operator who has not been 10-11 involved with the new product development

6

Redesigning the control panel to improve its visual impact and its 10-11 user friendliness and order new parts to build a second prototype

7

Conduct EMC and machinery directive compliance

12

8

Produce sales and promotional literature

9-12

9

Compile an instruction and maintenance manual

10-12

Figure 12: Industrialisation Stages and Schedule The total cost of completing these tasks is estimated to be of the order of 26 k Euro (including internal labour costs)

16.

Economic impact and improvement in competitive position

During the period of the application experiment, Blossom Design has experienced very strong price competition from its European competitors and more recently has seen one competitor exhibiting a new microcontroller based operating system at an exhibition in Cologne. This evidence of competitor technical progress indicates the importance of this project to Blossom Design operating in an evolving and innovative industry. Without this application experiment, they would experience a gradual worsening of their trading position, rather than their current expectation of maintaining market share in a growth industry. The universal recycling potential and the increased productivity performance offered by the improved recycling machine means that the performance of the improved product will match that now being offered by some competing products. This improvement in market competitiveness will maintain sales of the Dragon recycling machine in the short term and facilitate further refinement and development of the control system as the applications for this type of product increases, thereby leading to additional sales.

Page 28 of 311

29971

The added value provided for the customer includes: a) Increased productivity due to the improved control system, and the increased throughput. b) The ability to process a greater range of materials with the one machine. c) To be able to integrate the machine into upstream and downstream processes if required. These sales are occurring in a growing market for recycling products; this growth is partially a result of the increasing level of legislation throughout Europe and the USA in terms of recycling in general, and especially in terms of recycling packaging. Many industry sectors are also increasingly aware of the cost advantages to be gained from recycling waste materials and this also is stimulating the growth of this sector. This is especially true of the American market where the technological advantage of the machine will be especially valuable in providing a marketing advantage. The impact of the improved competitiveness of the universal recycling machine on Blossom Design’s sales is shown in Figure 14.

Sales Revenue Relative to 1999 figures % 200 Improved Product

150

Existing Product

100 50 0 1999 2000 2001 2002 2003

Year

Figure 14: Projected Sales Growth Sales relative to 1999 figures (%) The sales of the existing product are indicated as falling slightly in the future if improvements in competitive position are not implemented. The introduction of the improved Dragon Shredder product will correct this declining sales scenario as the improved technical performance provides sales stability. The projected sales growth is primarily in Europe, but based on sales of 6 units in the USA in the last 2½ years and potential in other markets, sales outside Europe are also expected to rise by 10-15% per annum. Erosion of profitability will also be reversed by the product improvements. The financial model for calculating the impact of the new products takes into account:

Page 29 of 311

29971

• the number of improved product units forecast to be sold vs. the old products • the planned selling price of the improved new products vs. the old products • the split between European and other world markets • the projected profit margin year by year for both the new and existing products The overall effect will be as shown in the forecasts in Figure 15.

Profit Forecast (1999 = 100%) 250 200 150

New Profitability Existing Profitability

100 50 0 1999 2000 2001 2002 2003

Figure 15: Profit Improvement forecast (1999 Profit = 100%)

Further potential financial improvement is thought to exist, owing to a likely reduction in the number of visits to customers’ premises, with a new product, which is adaptive to a degree. These savings have not been quantified here and are thought not to be large in relation to other improvements, at least in the short term. Payback Analysis: Based on a investment cost of 47 K Euro FUSE investment, and and assuming steady sales growth in the first year, the payback period for the application experiment is expected to be approximately 16 months. The additional industrialisation costs are estimated as 26 K Euro. Return on Investment over the first four years will be 600%, or 390% including the costs of industrialisation and launching the improved product.

Page 30 of 311

29971

17.

Target audience for dissemination throughout Europe

This application experiment demonstrates best practice in the specification processes, and the implications for the specification process itself on improvements in product design, project management, subcontractor selection and management, and in the process of testing adopted to deliver a functional prototype product for a hostile end environment. The primary industry sectors for this application experiment is the recycling industry. This sector is an increasingly expanding economic activity in the EC, and offers potential for the introduction of improved and more sophisticated microelectronics to raise the efficiency of the processes adopted in recycling However, the

target audience for the dissemination materials produced by Blossom Design includes small to medium sized enterprises (SMEs) in the following industrial sectors: Recycling - Prodcom code 37 Machinery - Prodcom code 29 Metal Fabrication - Prodcom code 28 Process Control equipment - Prodcom code 33

Page 31 of 311

Suggest Documents