↓ Skip to main content
Rillsoft AI
Connect Rillsoft

AI examples in Rillsoft Project

The right resource despite a site bottleneck

Is the branch crew enough for the track possession? Rail example: shortfall per role with and without the preferred team – the bottleneck is the site.

All people, companies and projects shown are synthetic Rillsoft sample data – from the shipped industry packages or, where stated, from a test project.

The question

Is the branch's own working group enough for the possession of track 1 (the period in which the track is closed) – and if not: is the skill missing, or only its binding to the site?

The data

Sample project track and turnout renewal at Rheinstadt station (a Rillsoft GmbH rail construction test project – not part of the shipped packages; the steps can be reproduced with any own project that carries a preferred team on a sub-project): 28 people in 13 roles across eleven disciplines, four working groups North A, South B, West and East. The project runs from October 2026 to February 2027 – construction phase 0 prepares, phase 1 closes track 1 from 24 Nov to 18 Dec 2026, phase 2 track 2 in January 2027. For the example phase 1 gets the working group North A as its preferred working group (in the product: preferred team) – the English project file does not carry it yet, the first step sets it: only its people count as capacity. All people and dates are synthetic.

The steps

  1. Set the preferred working group

    The AI opens the project, reads the project tree, sets North A as the preferred team of the sub-project construction phase 1 – possession of track 1 (one undo step) and opens its properties.

    Rillsoft Project, Gantt chart of the track and turnout renewal at Rheinstadt station, calendar weeks 44 to 3 (November 2026 to January 2027): construction phase 1 – track 1 closure highlighted with the tasks set up possession and safety measures, track removal track 1, formation improvement and drainage, track renewal and turnout installation track 1, overhead line adjustment track 1, tamping, grinding and acceptance measurement and the milestone track 1 ready for traffic; at the bottom the sub-project dialog, tab preferred team, with North A, South B, West and East on the left and North A as the assigned team on the right
    Sub-project construction phase 1 – track 1 closure: on the preferred team tab only North A of four working groups is assigned – the capacity of the possession is calculated from this group

    In the Gantt chart phase 1 is selected; the sub-project dialog shows on the preferred team tab the four working groups on the left and the assigned one on the right: North A.

    Construction phase 1 – track 1 closure runs from 24 Nov to 18 Dec 2026 with six tasks from setting up the possession to tamping, grinding and acceptance measurement, and the milestone “track 1 ready for traffic”. Preferred working group: North A.
  2. Capacity view by project and role

    The AI switches to the human resource capacity planning grouped by project and role, weekly scale, October to January, and reads the report.

    Rillsoft Project, human resource capacity planning by project and role, calendar weeks 41 to 1, week 1:2: row 3 construction phase 1 – track 1 closure with shortfall −224 (37 %) in red, below it 4.001 track superstructure – track worker −144 (43 %) with Kowalski and Hansen, 5.001 overhead line −40 (33 %) with Ludwig and Nowak, 7.001 civil works −40 (33 %) with Petrov and Gerdes; pink cells in calendar weeks 48, 49, 50 and 51 (track workers −32, −32, −48, −32); at the bottom the sub-project dialog still open with North A
    Capacity planning with preferred working group North A: phase 1 is short by 224 hours – three roles with two people each carry the gap

    Phase 1 shows −224 (37 %) in red. Below it the roles with a shortfall: track workers −144 (43 %), overhead line −40 (33 %), civil works −40 (33 %), each expanded with its two people from North A; the pink cells lie in calendar weeks 48 to 51.

    Construction phase 1 is short by 224 hours, 37 % of its demand of 600 hours. Track workers are short by 144 hours (43 %): 32 in calendar week 48, 32 in week 49, 48 in week 50, 32 in week 51. Overhead line is short by 40 hours (24 in week 50, 16 in week 51), civil works by 40 hours (24 in week 49, 16 in week 50). Only North A counts as capacity: two people per role.
  3. The bottleneck in the chart

    The AI opens the resource chart below the table for the row track workers; it draws this role only, per day.

    Rillsoft Project, the same human resource capacity planning with the row 4.001 track superstructure – track worker selected; at the bottom the resource chart of that row per day: green capacity line at 2, blue demand 4 with orange shortage from 25 November to 1 December and from 9 to 15 December, from 16 December demand 2 within capacity
    Resource chart of the track workers: track removal (25 Nov to 1 Dec) and track renewal with turnout installation (9 to 15 Dec) need four per day against a capacity of two – the orange bars are what North A cannot cover

    Per day, for this row only: the green capacity line stands at 2; during track removal (25 Nov to 1 Dec) and track renewal with turnout installation (9 to 15 Dec) the demand reaches 4 – orange shortage of 2 above the line; from 16 December the demand drops to 2 and stays within capacity.

    Track removal from 25 Nov to 1 Dec and track renewal and turnout installation from 9 to 15 Dec each need four track workers per day against a capacity of two. The gap is two per day – in hours 144 across weeks 48 to 51. Tamping, grinding and acceptance measurement from 16 to 18 Dec need two and are covered. The people of North A on this row are Kowalski and Hansen.
  4. Cross-check without the working group binding

    The AI removes the preferred working group from the sub-project – one undo step – and reads the same view again.

    Rillsoft Project, human resource capacity planning without preferred working group: construction phase 1 without shortfall, 4.001 track superstructure – track worker with Kowalski, Hansen, Aydin and Pohl, 5.001 overhead line with Ludwig, Nowak and Feld, 7.001 civil works with Petrov, Gerdes and Lang; at the bottom the resource chart of the track workers with the green capacity line at 4 and blue demand of at most 4 within it, no orange shortage
    Without the working group binding the capacity is sufficient: four track workers, three overhead line technicians, three civil works specialists – the shortfall was the site, not the skill

    The same capacity planning without red: track workers now count four people (plus Aydin and Pohl), overhead line three (plus Feld), civil works three (plus Lang). In the chart the green capacity line of the track workers now stands at 4 – exactly the peak demand, no orange shortage.

    Without the binding to North A phase 1 is covered: the capacity line of the track workers rises to 4 per day and meets the demand, the row shows a surplus in every week (+21.3, +32, +21.3, +48 hours in weeks 48 to 51); overhead line and civil works count three people each. The bottleneck is not the skill but the working group – Aydin and Pohl belong to working group West. One undo restores the preferred working group.

The result

The answer separates two questions that look the same in the capacity view: is the skill missing, or is it only missing in the preferred working group? For management: the possession of track 1 cannot be run with North A alone – 144 track-worker hours are missing across the four possession weeks, two people per day. With the two track workers from West it is covered. Whether the working group on the sub-project is opened or two people from West are assigned deliberately is the planner’s decision, not the filter’s. Without capacity per role both cases would be one red row; only demand against the working group’s supply shows that two track workers are missing – and where they are.

What runs behind it

  • rillsoft_project_file_open — File or session operation – not an undo step
  • rillsoft_project_tree_get — Read-only mode is enough
  • rillsoft_project_team_preference_set — Changes the project (one undo step)
  • rillsoft_property_panel_open — Read-only mode is enough
  • rillsoft_view_state_apply — View – one undo step, changes no project data
  • rillsoft_timescale_apply — View – one undo step, changes no project data
  • rillsoft_row_selection_apply — Read-only mode is enough
  • rillsoft_analysis_report_get — Read-only mode is enough
  • rillsoft_project_team_preference_remove — Changes the project (one undo step)
  • rillsoft_history_undo — File or session operation – not an undo step

Limits and variants

The preferred working group acts on the sub-project as a capacity filter: the capacity view counts only its people, demand stays the same. Removing it is a change to the project – one undo step, no file write; saving happens only on explicit request. The weekly hours per person refer to the open project, pro rata for people with several projects; the package carries no absences. Read-only mode is not enough for steps 1 and 4 (setting and removing the working group are changes); steps 2 and 3 run read-only once the view is set. Preferred teams in Rillsoft Project are described on the product page. Related prompts:

Reproduce it yourself

Get started
Industry package
AI/EN/Construction
Product version
Rillsoft Project 10 (Build 10.0.623)
Product language
English (test project rail construction)
Last verified
2026-09-27
Verification
Figures and statements verified on the named sample package against the named product build – MCP session, view and report; German and English package with identical figures.