Digital Transformation

Toyota's AI push to unify a fractured production language

Published Modified
2 min
Toyota to unify a fractured language of vehicle production

Toyota is deploying artificial intelligence to reconcile 45,000 internal terms into 5,000 common ones across planning, production and sales, aiming to cut translation steps by 30% and speed vehicle delivery by 2028.

Few carmakers illustrate the cost of institutional history quite like Toyota. The carmaker's plan, reported by Nikkei Asia, to standardise the language used in its specifications lists across planning, production and sales divisions is at this point, beyond a simple tidying exercise. The move is in essence, an attempt to dissolve a linguistic fault line that has run through the organisation for over seventy years.

The root of the problem stems from 1950 when Toyota split into separate sales and manufacturing companies. Each half went on to build its own systems for recording vehicle data, developing terminology in isolation from the other. The two entities merged again in 1982, by which point the divergent naming conventions they had accumulated were unable to remerge with them. The result, today, is an information architecture in which a single component can carry several names depending on who is describing it and at what stage of the process.

The scale of the mismatch is worth commenting on. A part the planning department might label "A1AC without audio system" could be recorded by the production system simply as "136S".

Multiply that kind of discrepancy across an entire vehicle's specification and the inefficiency compounds quickly.

The cost of being 'lost in translation'

Toyota's own analysis puts a number on that inefficiency. The end to end process, from vehicle planning through to delivery, comprises 560 steps. Of these, 195, or roughly 30% of the total, exist purely to translate terminology between departments, spanning some 800 separate business systems. Collectively, Toyota's employees spend approximately 310,000 hours a year on this conversion work alone, effort that adds no value to the vehicle itself.

Since 2021, Toyota has been building a corrective system known as OMUSVI, the Organised Master Unified System for Vehicle Information. Its ambition is to create a single integrated data layer spanning the full value chain, from planning to sales, with complete operational synergy targeted for 2028.

The AI component of the system is designed to automatically standardise terminology and enable real time sharing of vehicle information, so that, for instance, the sales division could monitor production progress directly rather than waiting on translated reports.

Mapping Toyota's translation tax

Toyota · OMUSVI

Seventy years of translation

How a 1950 corporate split left Toyota with two vocabularies for the same vehicle - and what it costs to bridge them.

1950
Sales and manufacturing split into separate companies
1982
Companies remerge; terminology stays divided
2021
OMUSVI development begins
2024
Koji Sato endorses the initiative publicly
2028
Full operational synergy targeted
71 years of divergent terminology One data layer targeted by 2028
560
steps from vehicle planning to delivery
195 (~30%)
of those steps exist purely to translate terminology
800
separate business systems spanned by the translation
310,000
hours spent annually converting terminology between departments

Same part, two names

A1AC without audio system
Example - Planning department label
136S
Example - Production system code
720 hrs
Time some suppliers spend each year manually calculating their own annual capacity allocations — a task OMUSVI is designed to remove by transmitting detailed specifications and quantities directly.

Consequences for suppliers

Studio portrait of a man in a dark suit, white shirt and blue tie wearing glasses against a light background.
Koji Sato, then-President, Toyota Motor Corporation

The implications, of course, extend beyond Toyota's own walls. The new system is intended to interface directly with annual production planning, automatically generating the required component volumes and transmitting that data straight to suppliers. At present, many suppliers must manually calculate their own annual capacity allocations, a task that for some companies consumes up to 720 hours a year.

Under the new arrangement, annual planning documents would instead carry detailed specifications and quantities for every component, removing much of that manual burden.

Internal resistance and the road to 2028

But the initiative has not proceeded without friction. Early implementation met resistance from departments concerned about the potential for job losses, which slowed its progress. 

Momentum picked up only after Koji Sato, then president of Toyota Motor Corporation, publicly endorsed the initiative in March 2024, lending it the executive weight needed to push through internal opposition.

Once fully deployed, the system is expected to meaningfully improve efficiency across Toyota's production and supply chain, translating fewer words into faster cars.