Co-Act
This project was a collaboration with Komatsu Forest AB. CoAct revolutionizes mechanic-customer interactions using data integration and 3D simulations. It addresses the loss of seasoned mechanics by preserving and transferring their knowledge. Combining big data with human expertise, it boosts operational efficiency and decision-making.
Project Information
Partner
Komatsu Ltd. & Komatsu Forest
Role
Critical Framing, Research through Design & UX
Team
Hanxiong Zhang, Lin Wang, Dide Sevincok
Context
CoAct is a cutting-edge tool designed to enhance the interaction between mechanics and customers, utilising sophisticated data integration and 3D simulations. It proactively addresses the impending departure of seasoned mechanics by safeguarding against knowledge attrition and promoting efficient knowledge transfer, ensuring preparedness and expertise in every service encounter.
What is the problem space?
The forestry sector in Sweden is a highly expertise-driven industry with significant economic influence. Companies like Komatsu Forest play a critical role through advanced machinery that powers mechanized harvesting. The maintenance and repair of this equipment require more than technical skills—field mechanics rely on tacit knowledge gained through experience.
Machine Operator
Caregiver
Harvester or Forwarder
Patient
Mechanic
Doctor
Max/ Forestry Machine Operator
Harvester
Oscar/ Komatsu Senior Mechanic
If you think about it, the operator is like a caregiver, the machine is the patient and the mechanic is a doctor.
This analogy helped us look at the problem area from a humanistic perspective. We opened up the idea of repair like a doctor would diagnose an illness. Which led us to the term "prognosis".
As we delved deeper into the process repair, we realised that it could be viewed as a dynamic process:
Before
Prognosis
During
Communication & Arrival
After
Fieldwork
Ethnographic Research
Field mechanics often engage in dynamic work environments where practical knowledge and human interaction are paramount. From workshops to on-site service work, the exchange of tacit knowledge through casual conversations and peer interactions is invaluable. Such environments not only enhance technical skills but also build a supportive network fostering innovative problem-solving.
Key quotes from the research
George/ Komatsu Senior Mechanic
Pointability
Standardisation of
a problem
'…Customers try to pinpoint the problem.'
'…Different customers describe the problems differently.’
'…The crane is not moving, they see the physical problem
but not the possible solution or spare parts.'
George/ Komatsu Senior Mechanic
'…If you can see the current data that causes the problem to the specific item can be a great help.'
'…The error in the cabin comes very very slow, the operator doesn't recognise until it comes but it could have been changed much earlier.'
' …The mechanic took apart the saw motor and he understood that the problem was not there at all.'
Prognosis tools
Decreasing back-forth by
prognostic troubleshooting
George/ Komatsu Senior Mechanic
' …If we had a 3D model with all the information running in real-time then we could do a diagnosis from our computers.
Digital Twin
Outcome
The operator reports the problem onsite
The mechanic can look at the digital- twin of the machine, while also viewing its health, issues and other data collected by the sensors and also reported by mechanics in the past
To make a prognosis, the tool utilises data logged by mechanics in the past; leveraging big data to predict the actual problem.
Knowledge transfer relies on formal training and on-the-job learning, making it difficult to preserve the nuanced insights from senior mechanics. The goal was to implement AI that integrates human experience with technology, ensuring engagement rather than isolation.
Reflection
One of the most critical UX challenges in this project was bridging the gap between human knowledge and machine intelligence — specifically, preserving the deep, often undocumented insights of experienced mechanics. Their know-how doesn’t live in databases; it lives in gestures, in routines, in conversations. Designing a system that respects and captures that tacit knowledge, while remaining intuitive and rugged enough for the field, pushed us to rethink how AI can support — rather than replace — human expertise.
A defining moment came during field research, when a mechanic told us how empowering it felt to have their insights valued in the design process. That moment reframed our approach: this wasn't just a technical challenge; it was a cultural one. We weren't just designing for usability, we were designing for trust, inclusion, and recognition.
Looking ahead, I see this project as part of a larger shift — one where human intuition and data-driven insight are not at odds, but deeply intertwined. Our goal is to build tools that amplify human capabilities, not obscure them. With the right scaffolding, AI can become a conduit for shared knowledge — connecting generations of expertise, supporting critical decision-making, and evolving alongside the people it serves.
Home