After a BSc degree in experimental physics and maths from UCD in 1967, I emigrated to England to work with the Marconi Company in Chelmsford, Essex. Marconi Training College gave me a good grounding in electronics before I worked on both digital and analogue circuitry as a design engineer in the line communications division. Digital Equipment Corporation (DEC) recruited me in February 1972 as a component engineer at its facility in Galway.
In 1972, the PDP-8 minicomputer was the mainstay of production and the Galway operation was very much still in a start-up mode. The key people were almost all US or UK transfers. The focus was on meeting production commitments with efficiency and cost control given second place.
It was an exciting place to work with multiple mini-crises to resolve. I got great job satisfaction solving many of these problems. Job descriptions were deliberately vague. This gave great latitude to individuals with the ability and ambition to expand their roles. When identifying a problem to a manager, you inevitably were given the problem to solve it yourself and when faced with difficult decisions you were told to ‘do the right thing’. All this was very empowering if you had the right attitude.
Over the next two years, many of the early deficiencies were brought under control. Key to this were hiring qualified and experienced engineers for the different operations in the facility and the in-house training of technicians. We also expanded production activity to include DEC printers and storage devices.
The opening of the first phase of Ballybrit in 1973 not only heralded a major expansion, but also a transition to more professional operations management. We now produced the PDP-11 family, which ranged from the small 11/05 to the large 11/70. The smaller computers were used mainly as process controllers while 11/70s were significant enterprise systems. The facility was divided into a ‘volume’ manufacturing area and a systems manufacturing area. The latter provided the customer-specific assembly and testing functions that the 11/70 needed. Each area had dedicated materials and product management resources. Support departments like engineering, finance and personnel provided services to both.
In 1974 I became the manufacturing engineering manager. This gave me responsibility for all the manufacturing processes in addition to incoming test and component engineering.
I developed a model with the finance department to select high margin products which would take advantage of Digital’s 0% tax on profits from products manufactured in Ireland. This model significantly improved corporate profit after tax and helped overcome opposition from design managers based in the US (powerful people in the Digital world).
We developed a much more professional introduction process for new products with minimal interruption to ongoing operations in the facility. Similar improvements were occurring in every department. For example, a focus on ‘just in time’ materials management had a big impact on stock holding.
After the announcement of the VAX 11/780 in October 1977, the senior management team in Galway realised that this new product was destined to be Digital’s flagship product for the next five to ten years. Nothing on its scale and complexity had been introduced in Galway before this. But Galway would have to introduce the system successfully – and as soon as possible – to fulfil its mandate to meet European demand.
They decided that they needed an experienced and dedicated manager to lead this project. At this time I was the manufacturing engineering manager responsible for new product introduction and technical support of the various production areas. So through this role I had the requisite experience. I was invited to take on the new challenge on a full-time basis.
It was agreed that I could select my own project team and, as the implementation plan progressed, people would be transferred from their old roles into new VAX production units. The project team members would then be managing their own groups and would grow into these roles gradually.
This transition would result in a vertically integrated VAX manufacturing unit under my management. A basic assembly and test (BAT) area would produce ‘modules’ – single circuit boards – and CPUs. The VAX modules were larger in size and more densely populated with integrated circuits than anything previously experienced. A final assembly and test (FAT) area would configure each system according to a customer’s order.
Over four months I developed a project plan through negotiations with Jerry Bourque, the US project manager, and with design engineering and manufacturing managers. Five separate facilities were contributing to the introduction of the VAX in the US – two each in Massachusetts and New Hampshire and one in Puerto Rico. We encountered resistance in some quarters to reproducing their work in Galway. Some of the Americans took a strong position against Galway making their products, expressing great scepticism that the west of Ireland could provide the technical resources and skills required. In several instances I invited the individuals concerned to come to Galway to witness its capability and to assess the calibre of the people there. They all went home as enthusiastic supporters.
In all previous new product introductions the US facilities had supplied Galway with kits of modules and other materials for the first month’s production. These facilitated debugging and reduced the learning curve. With the VAX, in contrast, customer demand in the US was so great that no such commitment was forthcoming for at least a year.
Despite the daunting task of managing the assembly and test of 20 different CPUs, I/O and memory modules and power supply units (which were significant products in themselves), we decided to forge ahead without kits. We would, however, be prudent with the ramp rate.
Several other innovations were agreed. We would eliminate the need for buffer stock between the BAT and FAT areas, thus shortening the process time. We also changed our procedure for configuring computers with multiple peripherals of the same type. Many standalone peripherals were manufactured in other DEC sites and imported to Galway to meet customer requirements. Before the VAX, each of these was tested as part of the full system. This involved removing their packaging and thus exposing them to cosmetic damage (scratches and bumps), which would require touch-up and finally re-boxing them for shipment to the customer. All of this hardware also put pressure on floor space – some systems would have 40 VDUs. We switched to a ‘dock merge’ strategy for the VAX. We stopped opening every box and took the small risk of shipping peripherals that had not been tested with their CPU.
We had an excellent training department in Galway managed by John Lewis. At an early stage in the plan he sent technical trainers to the US to study the functionality of the CPU modules and of the new VMS operating system. The goal was to teach the start-up technicians the theory in Galway and to follow up with practical training in the US.
The project plan and budget were finalised and approved and we moved into the implementation phase in May 1978. Our goal was to make the first European customer shipment of a VAX 11/780 in November.
The project team was then formed. It consisted of Enda Folan in the role of engineering and quality manager, Jimmy Laffey as BAT manager, Kevin O’Reilly as FAT manager and Gay Cooley as materials coordinator. Each spent some time with his counterparts in the US, getting first-hand experience with the product and processes. They formed vital relationships which enabled agreements, such as hands-on training for Galway technicians in the US.
Jimmy and Kevin selected and assigned the technicians and assemblers. They came on board in phases and their training plans were implemented.
Bills of material were set up and orders were placed. We had a commitment for the US to supply 20 modules to use as models in the assembly area and a further 20 fully functional modules for the test area. These had to be chased on an ongoing basis until we had the full complement. They then had to be kept up to date with ongoing engineering changes. Gay did a great job in managing this vital requirement.
Jimmy set up two new spaces in the BAT area. One of these was for cabinet construction. To reduce the cost of transporting empty cabinets by air, as heretofore, Digital took a ‘flat pack’ approach to the design of the VAX cabinet. We used pop rivets to assemble the various components into sturdy cabinets. The second space was divided into slots, each of which accommodated a CPU cabinet. We populated these with cable assemblies, power supplies, back panels and modules, and then ran diagnostics to isolate and fix faults.
John (arrowed) with the first complete VAX system produced in Galway and the assembly and test team that built it.
(Photograph courtesy of Enda Folan. Photographer unknown.)
Each completed machine that passed a quality inspection was immediately moved to a slot in the FAT area, instead of being sent to a stockroom. It was then assigned to an FAT technician who would ensure the configuration was complete, perform operating system tests and eliminate all bugs. To reduce infancy failures in the field, these systems would run continuously for several weeks before final shipment approval. Kevin managed the technicians and ensured that all tests were completed satisfactorily.
Digital developed a piece of proprietary software to automate the effects of updating changes to the project plans. It was like a primitive version of Microsoft Project. This application helped me enormously to control the project implementation. It had the ability to identify the critical path of the project. By not allowing critical tasks to slip, we were able to keep on track and meet target dates. I always got the appropriate positive response from the implementation team whenever any critical milestones were in jeopardy.
The first CPU from the BAT area was completed on schedule in September. This formed the core of the first customer shipment from the FAT area, which was delivered on schedule in November 1978, together with all its peripheral devices.
In addition to project team reviews that monitored the implementation tasks, we held monthly meetings for all the members of our growing VAX family. These enabled everybody to hear the status of the project at the same time and a good team spirit developed through our question and answer sessions. We slipped seamlessly from the start-up phase into ongoing production and gathered more people as demand for the VAX 11/780 grew. We were all driven by a continuous improvement attitude and by the desire to show what Digital Galway was capable of.
I managed this burgeoning group until September 1979. In the following month I moved to Salem, New Hampshire to manage VAX module testing and basic assembly at the DEC facility there. The main objectives for me were to test myself in a new environment and to manage a much larger group.
My contract in the US was for three years, but I received a request to return to Ireland after two. Frank McCabe had been appointed the general manager of manufacturing in Ireland. Many of our European customers were also potential suppliers and were applying pressure on Digital to give them sales opportunities for their components. Frank had therefore decided to set up an external resources group in Dublin to source across Europe a high percentage of the components used in the Irish facilities.
This operation started in November 1981. Ken Lynn was hired to take care of the commercial issues and I managed the component engineering aspect. It took several years to get this established. Finding component engineers with the appropriate skills and experience was difficult. After five years we had succeeded in sourcing sheet metal components in Ireland and Scotland, multi-layer printed circuit boards in the Netherlands and Austria, and many integrated circuits from the subsidiaries of US multinationals in Scotland.
By 1996 we managed the sourcing and contracting of commodities valued in excess of $300 million.
Last edit: October 2017
© John Eyres 2017