Shigeo Shingo was an industrial engineer whose consulting work with Toyota in the 1950s and beyond produced two major contributions to what became TPS: poka-yoke (mistake-proofing) and SMED (Single-Minute Exchange of Die, the systematic reduction of changeover time).
Poka-yoke and building quality in
Shingo's poka-yoke concept — designing processes and tools so that errors either cannot occur or are immediately detected — is one of the manufacturing foundations of build-integrity-in. The Poppendiecks argued that software development should build quality into the process rather than relying on end-stage inspection. In software terms, this means test-driven development, automated testing, and continuous integration: mechanisms that surface defects at the point of introduction rather than downstream, just as poka-yoke devices prevent or immediately flag errors on the production line.
SMED and small batches
Shingo's SMED methodology addressed the economic logic of large batch sizes: factories ran large batches because changeover times were long, making small batches uneconomical. By systematically reducing changeover time, SMED made small batches viable, enabling just-in-time production. This is the manufacturing logic behind the Poppendiecks' argument for small batches and fast delivery in software — deliver-as-fast-as-possible and the elimination of inventory waste in seven-wastes-of-software. When the cost of switching between tasks or deploying software is high, teams unconsciously batch work to amortize that cost; reducing that cost removes the structural incentive for large batches.
Shingo's contributions reach the Poppendiecks primarily through taiichi-ohno's TPS synthesis and through james-womack and daniel-jones's lean canon, which documented both poka-yoke and SMED as core lean tools.