Mountain bike configurator
From manual quotation to a Tacton configure-to-order system
/images/bike-configurator/hero.webp
Tacton configurator screen for selecting a mountain bike's gears and brakes, with live price and weight totals
- Client
- Cykelspecialisten
- Context
- Mass Customization, DTU
- Year
- 2024
- Role
- Process analysis and product modelling
- Team size
- Group 6
- Tools
- Tacton · Product Variant Master · CRC cards · Process mapping
- Contribution
- Team of 6, process flows, class diagram, PVM engineering view
An AS-IS and TO-BE analysis of a mountain bike manufacturer's specification process, together with a Product Variant Master and a configurator specification, in order to move the company from a manual quotation process to configure-to-order.
The problem
The specification process took roughly 3 weeks against a target of 1 week, and quote correctness sat at around 75%, which means that one in four quotes had to be corrected somewhere downstream. This is due to the process being manual and passing through five departments, where each handover reintroduces the possibility of an error, and where the same product knowledge is held in several heads rather than in one model.
What I owned
The work was split across the group, and I was responsible for the AS-IS and TO-BE process flows, the object-oriented analysis and the class diagram, and the Product Variant Master, where I specifically handled the engineering view and the assembly process.
Modelling the product
The Product Variant Master was built across the part, engineering and customer views, and the class model captures the components that actually drive the variance, including the Gearshift class with its three shifter types. Here the interesting part is the constraints rather than the components, since a combination shifter requires 4-finger brake grips, and this kind of rule is exactly what gets lost when a specification lives in someone's experience rather than in a model.
The TO-BE scenarios
We designed and scored several TO-BE scenarios rather than assuming that full automation was correct, and a partitioned design structure matrix was used in order to reorder the steps and break the circular dependencies in the existing process. Furthermore the requirement specification and the vendor selection criteria for the configurator were written against Tacton.
What I would change
It is important to note that the brakes logic was the weakest part of what we modelled, since the dependency between shifter type and grip configuration is more complicated in practice than the class diagram represents, and here I would spend more time with the people who actually specify bikes before committing to that structure.
/images/bike-configurator/crc-card.webp
CRC card for the Gearshift class, showing responsibilities, constraints and a price/weight table