Four questions behind a credible manufacturing commitment

In 2020, Aoritsune had 47 open manufacturing investigations older than ninety days in a defined register. The executive dashboard showed the count, but it did not distinguish missing evidence from a shortage of specialist time or an investigation whose question had been poorly framed. Production leaders wanted faster resolution. Quality colleagues were concerned that a closure objective would turn a capacity problem into pressure on the conclusion.

I asked the quality head to classify the work remaining and identify the decisions management could legitimately make. We found repeated dependence on a small group of technical specialists whose time was also committed to expansion projects. I approved ¥240 million of additional annual resource and deferred part of a non-essential process-improvement programme. The finance director challenged the recurring cost; I accepted that challenge and required a review of the workload rather than an assumption that the additional posts would remain necessary indefinitely.

We introduced monthly discussions about evidence, dependencies and investigation effectiveness. The quality organisation retained authority over conclusions. Executive reporting showed items reopened after an inadequate first explanation as well as items closed. Managers could therefore see why a lower count was not, by itself, a sufficient measure of improvement. We also recorded production and financial consequences separately so they did not become an implicit instruction about the technical answer.

After eighteen months, the older-than-ninety-day register had fallen from 47 to 13. Four closures had been reopened during the period and were shown explicitly in the review. The remaining cases had named evidence requirements and owners. I would begin with specialist demand across the whole company earlier next time. Our initial assessment focused too narrowly on the investigation team and missed the competing work that kept drawing essential expertise away.

Our review of selected hospital products identified eight materials whose supply depended on one qualified source. Purchasing colleagues had already located possible alternatives, and the first proposal described diversification as a procurement exercise with a short timetable. Technical and quality leaders explained that the work differed materially between materials. A common completion date would obscure those differences and could encourage the business to count an option before it was usable.

I approved a programme with separate decisions for each material and a budget that included qualification work, ongoing supplier oversight and the cost of maintaining supply from more than one source. The commercial team questioned paying a higher price while the existing source was still delivering. I asked them to compare the cost with the specific dependency we were trying to reduce, without presenting the alternative as protection against every possible interruption.

Six alternatives completed the required internal qualification process over twenty-four months. Two remained open because comparability questions had not been resolved to the quality team's satisfaction. We removed those two from the count of available options and revised the supply assessment accordingly. The programme increased annual material and oversight cost by approximately ¥180 million. We also introduced a periodic check that an approved alternative remained operationally available rather than becoming an unused file.

The enterprise gained six defined options and a clearer understanding of two unresolved dependencies. It did not gain immunity from shortages. My contribution was to keep the commercial case, technical evidence and reporting language aligned through the programme. I would now involve production planning earlier in the selection stage. A source that satisfies material requirements can still introduce scheduling or handling implications that change how useful the option will be in an actual disruption.

A site team brought the board a ¥7.2 billion expansion proposal intended to increase output across a group of established products. The engineering case was well developed, and several executives favoured approving the complete programme while contractor availability was favourable. The underlying model assumed that testing and review capacity would grow in proportion to production. Our operating data suggested that some of those activities had a different workload pattern and could become the limiting step.

I supported the need for additional capacity but did not support the proposed sequence. We divided the investment into a ¥1.1 billion enabling phase and a later decision on the main equipment and building work. The first phase funded a more detailed workload study, selected laboratory improvements and changes needed to observe the process accurately. This delayed the proposed construction start by nine months and required us to revise the commercial volume forecast.

The subsequent work showed that two product families consumed substantially more specialist review time than the initial model allowed. The revised proposal included that capability and a different equipment sequence. Its total planned expenditure became ¥7.6 billion, higher than the original request, while the expected ramp-up was less abrupt. I presented both changes to directors rather than describing the enabling phase only as a saving or a risk reduction.

The board gained a capital decision connected to the full operating path. Colleagues also learnt that a complete construction design does not establish a complete business case. I would improve the earliest discussion with the site team: asking them to reconsider after extensive engineering work made the change feel like a rejection of their effort. A shared constraint model before design commissioning would have made the later disagreement more productive and less costly.

During my COO tenure, a series of batches at one site was held while an unexpected analytical pattern was investigated. The products had committed dispatch dates, and the commercial team estimated that postponement would shift approximately ¥620 million of invoicing into a later period. Site colleagues were already working extended hours. The immediate management temptation was to ask for a date when the matter would be resolved, even though the investigation had not yet established a dependable basis for that answer.

I asked the responsible quality leader to explain the evidence needed for the next decision, without requiring a release date. The executive team then revised the dispatch forecast and authorised customer communication through the established channels. We funded additional specialist support and adjusted other site priorities. I made clear that these were management actions to support the investigation, not conditions attached to a preferred conclusion. The authorised quality process continued to determine disposition.

The hold lasted nineteen days. Some batches were subsequently released and others required further work; the public example does not disclose product or technical details. Invoicing shifted by ¥590 million, close to the earlier estimate, and the company incurred ¥46 million of additional investigation and scheduling cost. The financial effect remained in the operating review even after later dispatch recovered part of the timing difference. We did not present the eventual revenue as proof that the original concern had been excessive.

This episode changed how I ask for forecasts around unresolved quality matters. A management plan can contain ranges and decision points without turning uncertainty into a deadline for the evidence. If repeating the response, I would establish a single commercial communication owner earlier. Several teams initially requested separate updates, consuming specialist attention without improving the decision or what customers were told.

As chief supply officer at Nerukawa, I reviewed a portfolio whose lower-volume presentations accounted for 9% of revenue but 28% of scheduled changeovers in the studied product group. Operations proposed a rapid reduction in the number of variants. Commercial colleagues objected that several presentations served longstanding hospital purchasing arrangements and could not be treated as interchangeable simply because their sales were small. The first discussion therefore placed two incomplete measures against each other: changeover burden and current revenue.

I commissioned a product-by-product review with commercial, technical and quality teams. We considered the operating burden, customer dependence, feasible transitions and the specialist requirements attached to any proposed change. I did not authorise a blanket withdrawal. Of twenty-three variants examined, seven entered a managed discontinuation process, five were retained with revised planning arrangements and eleven remained unchanged pending better evidence or because their continuing role justified the burden.

The transitions took twenty months, longer than the initial operating proposal. We maintained additional inventory and incurred ¥95 million of transition cost. In the reviewed group, annual changeovers fell by 17%, while revenue associated with discontinued presentations had been approximately ¥310 million. The business tracked customer communication and unresolved supply obligations through the transition rather than closing the programme when the product codes were marked for removal.

The result was a smaller, better-understood source of complexity, not a claim that all low-volume products were uneconomic. It also gave the board a more credible explanation of why some apparently inefficient variants should remain. I would establish the customer-dependence assessment before ranking the products next time. Starting with an efficiency list made colleagues defend or attack individual items before we had agreed the evidence needed for a responsible portfolio decision.