Lean and Six Sigma tools merge to cut collision workshop activities by 30 percent

When a car arrives at a collision repair workshop after an accident, it triggers a surprisingly tangled chain of events. Insurers must assess damage and fault, qualified technicians produce appraisals to establish liabilities, and repair teams juggle parts, planning, painting and delivery while customers wait. Every deviation from the standard procedure injects variability into the process, and variability in service operations reliably translates into higher costs. A new study argues that much of this tangle can be untangled with a single, visually simple tool, and it reports the results of putting that idea to the test on the shop floor of a Portuguese automotive workshop.

The research, conducted at Caetano Auto Alfragide (CAA), a collision repair centre in the Salvador Caetano Group that has been Toyota’s official importer and distributor partner in Portugal since 1968, proposes an Integrated SIPOC-VSM Model, abbreviated IMSV. The model fuses two foundational improvement tools from the Lean and Six Sigma methodologies. The first is Value Stream Mapping, or VSM, a Lean technique introduced by Womack and Jones in their 1996 book Lean Thinking, which maps the flow of value through a production process from raw materials to delivery, exposing activities that add value and those that constitute waste. The second is SIPOC, a Six Sigma framework that organises a process at a high level into its Suppliers, Inputs, Processes, Outputs and Customers, defining project scope and aligning stakeholders before detailed analysis begins.

On paper, the two tools sound complementary, and they are often applied in sequence: SIPOC first to collect and organise the information needed to develop a value stream map. But the study’s authors observed a critical inefficiency in that sequential approach. The information described in the SIPOC model was largely repeated later in the VSM, creating procedural duplication of exactly the kind the tools are meant to eliminate. IMSV resolves this by embedding a service-adapted VSM, similar in style to the Makigami variant used for visual management in Japanese Lean practice, directly within the Process field of the SIPOC structure. The result is a single visual artifact that centralises suppliers, inputs, activities, actors, value classifications and proposed improvements in one collaborative model, eliminating the redundancy that arises when the tools are applied one after the other.

The service context matters technically. Traditional VSM was designed for manufacturing, where processes are tangible and measurable, and it struggles with the variability of tertiary-sector operations, where human interactions and unforeseen interruptions dominate. IMSV therefore adopts a Makigami-like mapping style in which key processes are defined broadly and then detailed activity by activity, often using coloured post-its on a physical board or a digital whiteboard. Each activity is tagged with the individual actors responsible, recorded through abbreviations in a legend, and classified by value type: Customer Value Added activities that directly contribute to the service the customer receives, Business Value Added activities that are operationally or regulatorily necessary but do not directly create customer value, and Waste, which is neither and therefore becomes a candidate for elimination, simplification or redesign.

Application of the model began, as Lean practice prescribes, with Gemba Walks, the Kaizen technique of going into the field and seeing processes with one’s own eyes. The research team combined direct non-participant observation with semi-structured interviews, internal process documentation and regular weekly development meetings, complemented by daily Kaizen meetings. Participants included the National Collision Director, the Workshop Manager, the Collision Team Manager and operational members of the collision team. Activities were included in the map only when they were recurrent, operationally relevant and explicitly validated by at least one process owner and one operational actor, a triangulation approach designed to reduce subjectivity. The model itself was built digitally using the MIRO software, allowing the multidisciplinary team to develop and refine the map collaboratively.

The case study focused on CAA’s collision department, which repairs sheet metal damage and paintwork on multi-brand vehicles, with Toyota as the flagship brand. The centre is small in infrastructure and staff, ranking 11th of the roughly 18 collision concessions in Portugal by annual turnover, yet it accounts for around 4 percent of CAA’s national revenues. Its business volume is heavily protocol-driven: 85 percent comes from protocol clients, split between 74 percent insurers and 11 percent fleet managers, with the remaining 15 percent from private customers and reconditioning. This insurance-dominated mix is precisely the environment where variability from claims handling, appraisals and total-loss assessments destabilises the repair production cycle.

Once the current-state map was complete, the team counted activities by responsible individual, so that an activity performed by three people counted as three procedures. This counting method exposes duplication directly. Improvements were then identified, structured and discussed with workshop managers, marked on the map with Kaizen Burst symbols indicating the Lean or Six Sigma tool to be applied, whether Kanban, poka-yoke, visual management, statistical process control or SMED. Responsibilities were reorganised to reduce unnecessary handoffs, shorten decision cycles and strengthen process ownership, particularly at the interfaces between technical, administrative and customer-facing functions. The validated improvements were implemented, monitored and recorded in a future-state model, which then becomes the standardised reference procedure for any employee needing to check the correct process.

The quantified results are substantial. Across key and indirect processes combined, the total number of identified activities fell from 71 in the current state to 50 in the future state, an absolute reduction of 21 activities and a relative reduction of 29.6 percent, roughly 30 percent. The arithmetic mean of the individual reductions across the seven key processes was 26.9 percent. Planning showed the largest single improvement, with activities cut by 45.5 percent, from 11 to 6. Delivery fell by 40 percent, from 10 to 6 activities, while Production and the other indirect processes each dropped by 33.3 percent. Insurance appraisal decreased by 16.7 percent and Parts/OR by 15.4 percent, while Reception remained unchanged. The study is careful to note that these figures represent reductions in the number of process activities rather than time- or defect-based performance metrics.

Workload redistribution proved equally significant. Counting activities by responsible actor, the collision manager’s assigned activities decreased from 34 to 19, a 44.1 percent reduction, while the Toyota service advisor’s fell from 14 to 10, a 28.6 percent reduction, and the workshop manager’s from 2 to 1. The mean reduction across the six responsible actors was 22.1 percent. The authors argue that this rebalancing addresses a gap left unaddressed by existing SIPOC-based hybrid frameworks, which have been proposed for machine learning data readiness, circular economy supply chains, open innovation and higher education quality assurance, but which remain largely analytical or conceptual rather than hands-on improvement models. Prior workshop studies have reported gains such as a 5 percent improvement in Toyota’s Fix-It-Right rate over ten months and an 11 percent improvement across key processes in a DMAIC-based programme, and a Toyota dealership study reduced express maintenance service time from 120.06 to 64 minutes, a 53 percent improvement per service cycle, but none offered a simple integrated model of this kind.

The study also frames the broader stakes. Automotive after-sales and insurance ecosystems are critical components of national transportation infrastructure, and inefficiencies within them generate cascading effects, including longer repair times, higher insurance premiums, resource waste and avoidable emissions. A low-complexity, data-light framework that organisations with limited Lean maturity can adopt supports service sector modernisation and sustainable competitiveness, while clearer role definition and reduced process fragmentation promise healthier working conditions and lower cognitive workload. The authors acknowledge key limitations: the model was validated in a single case, without external expert review, customer satisfaction measurement, statistical significance testing or benchmarking against standalone VSM, SIPOC or DMAIC, so its findings support analytical rather than statistical generalization. They propose multi-case replication across sectors such as healthcare, financial services and telecommunications, longitudinal implementation in which each future state becomes the next current state, and independent expert validation as the next steps to substantiate the model’s broader applicability.

Subject of Research: Proposal for an integrated SIPOC-VSM model: A case study in an automotive after-sales services workshop

Article Title: Proposal for an integrated SIPOC-VSM model: A case study in an automotive after-sales services workshop

Article References: Botas, D., Neves, A., Matos, F., & Godina, R. (2026). Proposal for an integrated SIPOC-VSM model: A case study in an automotive after-sales services workshop. Heliyon, 12(16), Article e45590. https://doi.org/10.1016/j.heliyon.2026.e45590

Image Credits: AI Generated

DOI: 10.1016/j.heliyon.2026.e45590

Keywords: Proposal, integrated, SIPOC-VSM, model, case, automotive, after-sales, services, workshop, scientific research