Real-time value stream mapping
The value stream map on the wall still shows a clean left-to-right flow. On the floor the same SKU leaves welding, waits in a side bay, crosses to Building B for wash, comes back for assembly, then loops to a second site for pack. Nobody on the sticky-note map owns that handling cost.
What real-time value stream mapping is
Real-time value stream mapping rebuilds the material journey from continuous location and process events, not from a one-day workshop drawing. Classic value stream mapping (VSM) captures a believed flow: process boxes, inventory triangles, and a lead-time ladder drawn from interviews and a few observations. Real-time VSM keeps measuring how each tagged SKU or lot actually moves across stations, buildings, and sites, then prices the manipulation against what engineering expected.
That matters when the flow is not linear. Many plants route parts through separate processes that branch, return, and cross yards. The wall map still looks sequential. The live path is a web. Handling between those hops is often the largest logistics cost in the margin, and it barely appears in a static VSM refresh done once a year.
MES events versus real movement
A MES records technology-flow steps: start weld, complete wash, book to assembly. Those events describe the intended process graph. They do not prove how far the lot traveled, how long it sat between books, or whether night shift took a different building-to-building path than morning. An RTLS stream fills the gaps between scans with real movement, dwell, and empty or loaded travel.
The useful comparison is conformance over time. Does live movement match the technology-flow description this shift? This week? After a layout change? When the paths diverge, you see either a documentation lag or a real workaround: a blocked aisle, a preferred side buffer, a second site hop that planning never costed. Twinzo analytics can slice the same SKU across days and weeks so a one-off Tuesday detour does not set policy, while a stable night-shift variant does. That is the same honesty used for unknown processes on the live floor, applied to the full value stream instead of a single side job.
Costing manipulation per SKU across halls and sites
Real-time VSM turns path length, handoffs, and dwell into a handling cost per SKU or family. Distance across one building is only the start. Multi-building and multi-site hops add yard moves, trailer waits, and double handling at each door. Pair those minutes and meters (feet) with labor or fleet rates and the plant can say what manipulation actually costs versus the expectation baked into the process sheet or the annual VSM.
Expectation comes from the designed route: station sequence, standard travel, planned buffers. Reality comes from tagged movement on the twin used for internal logistics optimization. The gap is margin. Extra loops, return trips to a previous hall, and long WIP waits between non-linear processes show up as cost that speed projects never touch if they only optimize cycle time inside each cell. Spaghetti and heat maps make the waste visible. The SKU cost roll-up makes it negotiable in a relayout meeting.
Relayout, speed, and AGV or AMR decisions
Relayout calculations need more than a cleaner drawing. They need proof that moving a wash cell closer to welding, or staging a satellite buffer by the pack hall, cuts handling cost enough to pay for the civil work. Real-time VSM supplies before-and-after path cost on the same SKUs. Teams compare lead time and manipulation margin, not only walking intuition from one gemba tour.
The same evidence supports AGV and AMR decisions. Stable, repeated corridors with high manipulation cost are automation candidates. Chaotic, low-frequency hops across sites may stay on manned trucks. If the live stream shows night shift already running a tighter path than the map assumes, automate that path, not the outdated wall version. Throughput and bottleneck talks then rest on measured travel and wait, not on the process box sequence alone.
How teams run a live VSM loop
1. Pick SKU families with heavy handling - Multi-process, multi-building, or multi-site flows first. Tag the lots or returnable carriers that actually travel the hops.
2. Capture the expected technology flow - Station sequence and planned moves from MES, process sheets, or the last classic VSM. That is the baseline to beat.
3. Measure real paths over shifts, days, and weeks - Compare live movement to the expected graph. Flag stable variants by shift and ignore one-off detours until they repeat.
4. Price manipulation and test layout or automation options - Roll distance, dwell, and handoffs into cost per SKU. Use the gap to rank relayout moves and AGV or AMR corridors, then re-measure after the change.
What a matched versus drifted stream looks like
A typical moment: a housing family is supposed to run weld, wash, assemble, pack in one hall. Live paths for three weeks show 40% of night lots detour to Building C for a temporary overflow wash, adding about 220 m (about 720 ft) and a trailer wait each time. MES still books the same four steps. The real-time VSM prices the night variant above the day path by a clear handling premium. Engineering either restores wash capacity in the main hall or updates the official flow and staffs the hop on purpose. Margins stop assuming a linear map the floor abandoned months ago.
When live movement matches the technology description across weeks, the classic VSM was honest and you can invest in speed inside the cells. When it drifts by shift or by day of week, fix the flow geography before buying more station capacity. Related WIP search and put-aside capital sit under search and find with RTLS. For how RTLS works on the floor, see what RTLS is. A broader logistics analysis path is in logistics analysis with a digital twin. For when continuous location helps versus when a workshop map is enough, see when RTLS helps, and when it does not.
Map the cost the wall chart missed
If manipulation across buildings and sites still hides inside a linear VSM, instrument the moving lots and compare live paths to the MES story until relayout and AGV choices rest on measured handling cost. The digital twin pays when someone changes the geography, not when the dashboard only redraws the old sticky notes. Get in touch if you want to walk the same real-time value stream on your own facility model.