Repository navigation
Two implements on the same strip (e.g. front tank + seeder) are mapped to AOG separately #137
Description
@richardklasens, you asked about this earlier. Here is how the TC behaves with two implements on the same strip today, and what it would take to support it.
Related to #132 and #136. Not planned yet: no machine to test with. This records how the TC behaves today and what a fix would need.
Problem
Two ISOBUS implements working the same strip, for example a 6 m front tank and a 6 m seeder behind the tractor, are two TC clients with two DDOPs. The TC handles each one on its own:
- Demand: PGN 229 goes to every connected implement, and implement section i follows AOG section i (after that implement's own
groupOverlappingSectionsmapping from Basic multi-boom support; collapse identical/overlapping sections #131). Nothing links one implement's sections to the other's. - Heartbeat: one
0xF0is sent per implement, each with its own section count. AOG has one tool, so with two implements it gets two different reports every 100 ms. This is today's behaviour ondevelop. - Grouping (Basic multi-boom support; collapse identical/overlapping sections #131) and the planned covering rule (Grouping: a section covering several others should follow their OR (e.g. full-width fertiliser over planter rows) #136) only look within one implement.
| Front implement | Rear implement | Result today |
|---|---|---|
| 1 × 6 m | 1 × 6 m (or collapsed to 1) | Works: both follow AOG section 1. Two heartbeats, both "1 section", whose on/off can differ. |
| 2 × 3 m | 2 × 3 m | Works: same index on both. |
| 1 × 6 m | 6 × 1 m | Wrong: the front implement follows AOG section 1 only, so it switches off while sections 2–6 are still seeding. It should follow OR(1–6). |
Note that many front tanks aren't a second implement at all. They are driven through the seeder's ECU and appear as a Bin in the seeder's DDOP (the Amazone Precea DDOP from #35 has seed, fertiliser and microgranulate as Bins under one Function). Those are handled by the implement itself and aren't affected by this issue.
What a fix would need
- One AOG section map across all connected implements, not one per implement.
- Common coordinates. Each DDOP's offsets are relative to its own connector, and a front hitch and a rear hitch differ. Y offsets have to be measured from the tractor centreline before they can be compared. X can't be compared without knowing where each product actually lands; a front tank's product often lands at the seeder coulters, not at the tank.
- Same strip and covering rules across implements: equal strips share an AOG section, and a covering section follows the OR of what it covers (Grouping: a section covering several others should follow their OR (e.g. full-width fertiliser over planter rows) #136).
- One combined heartbeat reporting AOG's sections, on while any member section is on.
- A rule for implements that can't be matched (unknown geometry, partial overlap): keep the per-implement behaviour and log why.
Needed before building
- A real two-implement setup (front tank + seeder, or similar) with DDOPs from both.
- TC logs from the Basic multi-boom support; collapse identical/overlapping sections #131 build showing each implement's reported Offset Y and width.
- Knowing how AOG should be set up for it (one tool width, section count).
Activity