From Radiators to Road maps: Control Systems & Agile Product Development

Posted on March 5, 2025

Next week, I’m heading to London for Mind the Product! I’m excited to dive into product management inspiration, talks, and connect with fellow PMs.

As I get ready for MTP, I’ve been reflecting on my work in Agile product development. Specifically, I keep coming back to control systems as a powerful way to slice value increments. Mind you I am working for a company named AMCS - Advanced Manufacturing Control System.

Imagine this: Input -> Control System -> Output. Think of a thermostat – it senses, adjusts, and maintains the temperature. 🌡️ In my world of optimizing planning and scheduling for heavy-duty trucks, our planners and dispatchers are essentially controllers in a closed-loop system. They receive input, use sensors (decision support tools), and generate output (decisions). Which feeds into the system control system over and over again.

Why does this matter? Because it helps us break down complex systems into manageable value increments.

Let’s take an example: automating 3rd party transport supplier load balancing with internal transport capacity.

Input: Transport jobs, supplier pricing, internal capacity. Controller: Our transport planner. Sensor: Capacity consequence calculator. Actuator: Actioned decisions leading to a new plan. Output: Supplier cost, internal capacity adjustments.

By viewing it through this lens, we can deliver value in stages:

Enhance the input (e.g., visualize jobs on a map). Improve the actuator (e.g., simplify supplier switching). Strengthen the sensor (e.g., provide cost consequence analysis).

This approach also guides system development from scratch in sequenced value increments:

And when we aim to automate the controller (hello, AI!), the importance of accurate input and robust sensors becomes crystal clear. This is essential for building effective AI agents in transport management.

Looking forward to discussing these ideas and more at MTP! See you there?