The Third Way of DevOps: Continuous Learning and Docker as Lab Equipment

The Third Way of DevOps completes the full DevOps cycle and is often called “Continuous Learning,” emphasizing a culture of continuous experimentation and learning across all organizational activities.

Building on the First Way (left‑to‑right flow and systems thinking) and the Second Way (right‑to‑left feedback loops), the Third Way introduces a complete loop that ties the previous two ways together through a rigorous learning process.

The Third Way draws on Kaizen and the Plan‑Do‑Study‑Act (PDSA) cycle popularized by Edward Deming, treating every action as a small experiment grounded in the scientific method.

References to the Toyota Production System (TPS) highlight its Lean origins; Spear and Brown’s Harvard Business Review article and Mike Rother’s work on Kata illustrate how TPS applied scientific methods and continuous experimentation to manufacturing.

Docker is presented as the modern “lab equipment” that provides reliable, reproducible environments for these experiments, allowing teams to quickly spin up containers that encapsulate required tools.

A financial institution using Docker‑based Container‑as‑a‑Service enabled over 100 data scientists to match analysis tools to data sets in minutes, cutting experiment setup time from days to hours and allowing multiple tool trials per project.

In a personal example, the author and his 12‑year‑old son built an R‑based Docker image containing baseball statistics packages, demonstrating how containers can instantly provide access to extensive datasets for analytics.

Docker and the Third Way

The article concludes by urging organizations to adopt the Third Way and use Docker to dramatically reduce the mean‑time‑to‑experiment, thereby accelerating software delivery and service improvement.