Transport planning, routing & dispatch
Plan loads and routes and dispatch fleet in TMS/dispatch tools.
20 project-authored proposals · taxonomy v1.1 · 0 new benchmark runs. Verifiers and partial-credit rules below are designs, not executed results.
- Typical applications
- TMS · route planners · fleet dispatch
- Class GUI-need label
HY· Hybrid
Task proposals · 20
Expand to inspect the complete original wording and proposed checks. Original proposal text remains English in both interfaces.
R05.09.01.03-T01 Plan capacity- and window-feasible routes and fix optimizer break violations
Original instruction
Route4Me (sandbox) has 22 delivery stops imported from input/stops_1015.csv with time windows and pallet counts. Plan routes for 3 trucks (10 pallets each, depot DC-East, 07:00-17:00). The optimizer ignores the 30-minute lunch break required between 11:30 and 13:00; drag stops between or within routes until every route has a feasible break slot.
- Platforms
web· Any OS (web)- Applications
- Route4Me
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
drag· Drag & dropnav· Pan/zoom/scrollclick· Point & click- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Exported route plans recomputed with a fixed OSRM matrix: all 22 stops served once, capacity <= 10, windows met, each route has a 30-min idle gap in 11:30-13:00; total miles <= reference optimum +5%. Agent has no API key; plan revisions logged by UI.
Proposed partial credit
30% all stops assigned within capacity; 30% windows met; 25% break feasibility; 15% miles within 1.05x reference optimum.
R05.09.01.03-T02 Plan a day with driver shifts and liftgate skills
Original instruction
In OptimoRoute (test account), plan tomorrow's 41 orders from input/orders_1016.xlsx for 5 drivers with shifts from input/drivers.xlsx. Orders tagged 'LIFTGATE' must go to the two liftgate trucks; driver Ortiz ends at 14:00. Plan, then lock routes and export the plan to output/plan.csv.
- Platforms
web· Any OS (web)- Applications
- OptimoRoute · Excel
- Interfaces
brow· Browser- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicktype· Text & formsmulti· Cross-app- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
output/plan.csv and account state: every order assigned or explicitly unscheduled with reason; liftgate orders only on liftgate vehicles; Ortiz route ends <= 14:00; shift windows respected; routes locked. Unscheduled count <= reference.
Proposed partial credit
30% all orders imported correctly; 30% skill constraint; 20% shift constraints; 20% locked and exported.
R05.09.01.03-T03 Assign same-day tasks to drivers on a dispatch map by zone
Original instruction
Onfleet dispatcher dashboard (sandbox org) shows 15 unassigned same-day tasks and 4 on-duty drivers. Assign tasks by dragging them onto drivers so that each driver only serves tasks inside their team's hub zone shown on the map, no driver exceeds 5 tasks, and tasks with 'before 12:00' windows are ordered first in each route.
- Platforms
web· Any OS (web)- Applications
- Onfleet
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
drag· Drag & dropvis· Visual judgmentnav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Onfleet sandbox task/worker state: every task assigned once; task coordinates inside the assigned driver's hub polygon; <= 5 per driver; pre-noon tasks precede others in each route order. Onfleet API keys are withheld from the agent, so assignments can only be made in the dispatcher dashboard.
Proposed partial credit
35% all tasks assigned; 35% zone compliance; 15% load cap; 15% ordering.
R05.09.01.03-T04 Draw a road-closure avoidance zone and re-route one truck
Original instruction
In the route planner (Route4Me sandbox), truck 2's route crosses a closure described in input/closure_notice.pdf (map image with the closed segment). Draw an avoidance zone polygon covering the closed segment and its two intersections, re-optimize only truck 2, and send updated ETAs to its affected stops via the sandbox notification feature.
- Platforms
web· Any OS (web)- Applications
- Route4Me
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
path· Continuous pathvis· Visual judgmentnav· Pan/zoom/scrollclick· Point & click- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Avoidance zone polygon covers the closed segment geometry (buffer test) and does not cover the parallel open road; truck 2 polyline has 0 intersections with the closure; other routes unchanged; notification log has entries for stops with ETA change > 10 min. Route4Me API keys are withheld from the agent, so the zone and re-optimization must come from the web planner.
Proposed partial credit
35% polygon covers closure without over-blocking; 35% truck 2 re-routed; 15% other routes untouched; 15% notifications sent.
R05.09.01.03-T05 Load a 40ft container to stacking and weight rules
Original instruction
In EasyCargo (web), plan a 40ft high-cube container for the 26 carton types in input/packing_list.xlsx. Respect 'no stack' and 'this side up' flags, max gross 26,500 kg, keep the center of gravity within 1 m of the container center, and reach at least 88% volume utilization. Export the load plan PDF and XLSX to output/.
- Platforms
web· Any OS (web)- Applications
- EasyCargo · Excel
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
drag· Drag & dropgeom· Precise geometrynav· Pan/zoom/scroll- Work product
WP04· 3D/spatial model- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Parse exported XLSX placements: all cartons placed, no box overlaps, no-stack items have nothing above, orientation flags respected, weight <= limit, CoG offset <= 1 m, utilization >= 88%. Manual moves are required: auto-load is seeded to land at about 82%. EasyCargo API keys are withheld from the account, so plans can only be built in the web planner.
Proposed partial credit
25% all cartons placed legally; 25% stacking/orientation; 20% weight and CoG; 30% utilization target met (linear credit from 0.82 to 0.88).
R05.09.01.03-T06 Arrange pallets in a 53ft trailer within axle limits
Original instruction
In Cube-IQ on Windows, place the 22 pallets from input/pallets.csv into a 53ft trailer so that drive-axle load <= 34,000 lb and trailer-axle load <= 34,000 lb with tandems at position 'standard', heavy pallets on the floor, and stops unloaded in reverse order (stop 3 at the rear). Save the load and export the axle report.
- Platforms
win· Windows- Applications
- Cube-IQ
- Interfaces
desk· Desktop GUI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G3· GUI-only- Interaction primitives
drag· Drag & dropgeom· Precise geometryvis· Visual judgment- Work product
WP04· 3D/spatial model- Conditions
licensed· Licensed pro app- Proposed difficulty
- 4 / 5
Proposed verifier
Exported placement and axle report recomputed by beam-moment model: both axle loads within limits; stop order respected front-to-back; no stacked pallet heavier than the one below; all 22 placed. Cube-IQ's batch/API interface is not licensed on the VM and command-line launch is blocked, so the load must be built and saved in the GUI.
Proposed partial credit
25% all pallets placed; 35% axle limits met; 25% stop sequence; 15% stacking rule.
R05.09.01.03-T07 Tender four loads to the cheapest carrier meeting transit time
Original instruction
In Oracle Transportation Management (training instance), four planned shipments await tender. For each, select the carrier with the lowest total rate whose transit days meet the delivery date, using the rate offerings in the system and the carrier scorecard exclusions in input/excluded_carriers.txt. Tender and record acceptance from the simulator.
- Platforms
web· Any OS (web)- Applications
- Oracle OTM
- Interfaces
brow· Browser- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scrolltype· Text & forms- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
OTM shipment tables: each shipment tendered to the key carrier (min cost, on time, not excluded); tender status accepted; no shipment tendered twice. Integration user disabled for agent.
Proposed partial credit
80% correct carriers tendered (pro rata over 4 shipments); 20% acceptances recorded and no duplicates.
R05.09.01.03-T08 Consolidate freight units into freight orders in SAP TM
Original instruction
In SAP TM (sandbox, SAP GUI/NWBC), consolidate the 6 freight units from forwarding orders in input/fwo_list.xlsx into 2 freight orders by destination region (WEST/EAST), use vehicle resource TRUCK-24T, assign carrier per input/carrier_award.xlsx, and set both freight orders to 'Ready for execution'. Write their numbers to output/fo.txt.
- Platforms
win· Windows- Applications
- SAP TM · SAP GUI
- Interfaces
desk· Desktop GUI- GUI need · task setting
G3· GUI-only- GUI need · ordinary access
G3· GUI-only- Interaction primitives
click· Point & clickdrag· Drag & dropnav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
licensed· Licensed pro app- Proposed difficulty
- 4 / 5
Proposed verifier
TM tables: exactly 2 FOs, each containing the freight units of one region, resource TRUCK-24T, carrier per award, execution status ready; numbers match output/fo.txt. SAP GUI scripting is disabled server-side and the agent user has no RFC or OData service authorization.
Proposed partial credit
40% correct FU-to-FO grouping; 20% resource; 25% carrier; 15% status and numbers.
R05.09.01.03-T09 Compute practical truck miles for lanes with a hazmat profile
Original instruction
In PC*MILER (Windows) compute practical truck miles for the 8 origin-destination lanes in input/lanes.xlsx using routing profile 'Hazmat-Flammable', 53ft trailer, toll discouraged. Enter miles and drive time per lane in output/lanes_miles.xlsx and save the report PDF.
- Platforms
win· Windows- Applications
- PC*MILER · Excel
- Interfaces
desk· Desktop GUI- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicktype· Text & formsmulti· Cross-app- Work product
WP02· Records/table- Conditions
licensed· Licensed pro app- Proposed difficulty
- 2 / 5
Proposed verifier
Grader reruns lanes through the same PC*MILER build with the profile: miles within 0.5% and times within 2 min per lane; a non-hazmat profile result differs on 3 seeded lanes and is detected.
Proposed partial credit
70% lane miles correct (pro rata over 8 lanes); 20% drive times; 10% report saved.
R05.09.01.03-T10 Dispatch an urgent pickup in a live fleet simulation
Original instruction
A fleet dispatch console (training simulator, 4x real time) shows 9 moving trucks. An urgent pickup arrives at 10:42 sim time. Within 3 sim-minutes, assign it to the truck that can reach it first without breaking any committed delivery window, insert it into that truck's route, and confirm the dispatch.
- Platforms
web· Any OS (web)- Applications
- fleet dispatch console
- Interfaces
brow· Browserhmi· Device/HMI- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
rt· Real-time controldrag· Drag & dropvis· Visual judgmentclick· Point & click- Work product
WP08· App/system state- Conditions
interrupt· Interruptions- Proposed difficulty
- 4 / 5
Proposed verifier
Simulator log: assignment timestamp <= 10:45 sim; chosen truck equals key computed from positions at assignment time; all committed windows still met after insertion. Simulator exposes only the UI.
Proposed partial credit
40% correct truck; 30% inserted without window violations; 30% within the time limit (linear decay to 6 sim-min).
R05.09.01.03-T11 Build an ordered technician route and export it
Original instruction
In Circuit Route Planner (web), create a route for technician Lee starting and ending at the depot in input/depot.txt with the 9 service addresses in input/jobs.csv, one of which must be first (priority). Optimize the rest, start at 08:00 with 20-min stops, and export the route to output/route.csv.
- Platforms
web· Any OS (web)- Applications
- Circuit Route Planner
- Interfaces
brow· Browser- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicktype· Text & forms- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
output/route.csv: 9 stops, priority stop first, sequence total distance (fixed OSRM matrix) <= best-order distance +3%, ETAs consistent with 20-min service times.
Proposed partial credit
30% all stops geocoded right; 25% priority first; 35% near-optimal order; 10% export.
R05.09.01.03-T12 Draw dock geofences for customer sites on a fleet map
Original instruction
In Samsara (sandbox org), create polygon geofence addresses for the 5 customer sites in input/sites/*.pdf; each PDF shows a site plan with the truck dock marked. Draw each polygon around the dock yard only (not the whole parcel), name it per the PDF, and tag it 'customer-dock'.
- Platforms
web· Any OS (web)- Applications
- Samsara
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
path· Continuous pathvis· Visual judgmentnav· Pan/zoom/scroll- Work product
WP10· Annotations- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Address geofence polygons: IoU >= 0.6 with reference dock-yard polygons and area <= 1.5x reference; names and tags exact. No API token is issued to the agent, so geofences can only be drawn in the Samsara web dashboard.
Proposed partial credit
60% polygon IoU (pro rata over 5 sites); 25% names; 15% tags.
R05.09.01.03-T13 Assign loads to drivers within hours-of-service limits
Original instruction
The Motive (KeepTruckin) fleet dashboard (sandbox) shows 4 drivers with remaining HOS drive and duty hours. Assign the 6 loads in input/loads_1017.csv (estimated drive hours, pickup windows) so no driver exceeds remaining 11h drive or 14h duty, each load is covered, and deadhead miles are minimized. Record assignments in the dispatch module.
- Platforms
web· Any OS (web)- Applications
- Motive
- Interfaces
brow· Browser- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicknav· Pan/zoom/scrollmulti· Cross-app- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Dispatch records: every load assigned once; HOS feasibility checked from seeded logs; deadhead miles <= key optimum +10%; driver assignments match one of the feasible optimal sets.
Proposed partial credit
30% all loads assigned; 40% HOS feasibility; 30% deadhead within 1.1x key optimum.
R05.09.01.03-T14 Classify and book an LTL shipment by density
Original instruction
In the TMS (MercuryGate training tenant), book the LTL shipment in input/ltl_request.pdf (4 pallets with dims and weights). Compute density to set freight class per the density table in input/nmfc_density.pdf, enter accessorials (liftgate at delivery, appointment), rate-shop, and book the lowest-cost carrier. Do not tender for pickup.
- Platforms
web· Any OS (web)- Applications
- MercuryGate TMS
- Interfaces
brow· Browser- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
type· Text & formsclick· Point & clickmulti· Cross-app- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Shipment record: freight class per pallet equals key from density; accessorials set; selected carrier equals min-cost rate quote; status Booked, not Tendered.
Proposed partial credit
40% freight class correct; 20% accessorials; 30% cheapest carrier; 10% stopped before tender.
R05.09.01.03-T15 Plan paired pickup-and-delivery routes with time windows
Original instruction
In Routific (test account), plan 18 pickup-delivery pairs from input/pd_pairs.csv for 4 vans (cargo 12 m3) with time windows. Each delivery must follow its pickup on the same van, perishable pairs must be delivered within 90 minutes of pickup, and no van exceeds 9 hours. Manually fix any pair the optimizer leaves unserved.
- Platforms
web· Any OS (web)- Applications
- Routific
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
drag· Drag & dropclick· Point & clicknav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
long· Long-horizon- Proposed difficulty
- 5 / 5
Proposed verifier
Exported routes checked: pairing and precedence, ride-time <= 90 min for perishables, capacity profile <= 12 m3 at all times, shift <= 9 h, windows met; total time <= reference +7%. Seeded instance makes optimizer drop 2 pairs. Routific API keys are withheld from the test account, so the plan must be built and fixed in the web planner.
Proposed partial credit
25% all pairs served; 25% precedence and same-van; 20% perishable ride time; 30% capacity, shifts and total time within 1.07x reference.
R05.09.01.03-T16 Complete a delivery route in a driver app with proof of delivery
Original instruction
On an iOS simulator with the Onfleet Driver app (sandbox driver), start the assigned route of 5 tasks. For tasks 1-4 complete with a signature and the photo provided in the simulator photo library; task 5's recipient is absent: fail it with reason 'Recipient not available' and a note. Follow the in-app task order.
- Platforms
ios· iOS/iPadOS- Applications
- Onfleet Driver
- Interfaces
mob· Mobile touch- GUI need · task setting
G3· GUI-only- GUI need · ordinary access
G3· GUI-only- Interaction primitives
touch· Touch gesturespath· Continuous pathtype· Text & forms- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 2 / 5
Proposed verifier
Sandbox task records: tasks 1-4 completed with signature image and photo attachment, completion order matches route; task 5 failed with reason and note. Driver token not exposed; app is the only client.
Proposed partial credit
50% four completions with POD; 30% failed task with reason and note; 20% correct order.
R05.09.01.03-T17 Control: solve the T01 routing instance with a script
Original instruction
Control pair of T01. Using the same 22 stops (input/stops_1015.csv) and the distance matrix input/osrm_matrix.json, write a script (e.g. OR-Tools) that produces 3 routes satisfying capacity, windows and the 11:30-13:00 30-minute break, and write output/routes.csv (truck, seq, stop_id, eta).
- Platforms
lin· Linuxmac· macOSwin· Windows- Applications
- Python · OR-Tools
- Interfaces
term· Terminal- GUI need · task setting
CL· CLI-sufficient- GUI need · ordinary access
CL· CLI-sufficient- Interaction primitives
code· Code & commands- Work product
WP06· Code/engineering- Conditions
- None declared
- Proposed difficulty
- 3 / 5
Proposed verifier
Same feasibility checker as T01 run on routes.csv; distance <= reference +5%; script reruns deterministically within 60 s.
Proposed partial credit
30% coverage and capacity; 30% windows; 25% breaks; 15% distance within 1.05x reference.
R05.09.01.03-T18 Schedule field work orders on a drag-and-drop schedule board
Original instruction
In Dynamics 365 Field Service (trial), schedule tomorrow's 12 work orders on the schedule board by dragging them to technicians. Respect required characteristics (e.g. 'Gas certified' rating >= 3), territories, customer time promises and 30-min travel buffers, and keep each technician <= 8 booked hours.
- Platforms
web· Any OS (web)- Applications
- Dynamics 365 Field Service
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
drag· Drag & dropnav· Pan/zoom/scrollhover· Hover & context- Work product
WP08· App/system state- Conditions
- None declared
- Proposed difficulty
- 4 / 5
Proposed verifier
Bookable resource bookings: all 12 scheduled; characteristic, territory and promised-window constraints satisfied; booked hours <= 8; travel gaps >= 30 min. Agent has no API token for Dataverse, and bookings must carry the schedule-board creation source.
Proposed partial credit
30% all booked; 30% skills and territory; 25% time promises; 15% capacity and travel gaps.
R05.09.01.03-T19 Book an ocean container on a sailing that meets cutoff
Original instruction
On the ocean carrier booking portal (sandbox), book one 40HC from Shanghai to Long Beach for the cargo in input/booking_req.pdf. Choose the earliest sailing whose cargo cutoff is after the cargo-ready date + 2 days and whose arrival meets the required date; enter commodity, HS code, weight and VGM method. Stop at 'Submit booking'.
- Platforms
web· Any OS (web)- Applications
- ocean carrier booking portal
- Interfaces
brow· Browser- GUI need · task setting
HY· Hybrid- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
click· Point & clicktype· Text & formsnav· Pan/zoom/scroll- Work product
WP08· App/system state- Conditions
risk· Irreversible/risky- Proposed difficulty
- 3 / 5
Proposed verifier
Draft booking record: selected voyage equals key; equipment 40HC x1; commodity/HS/weight/VGM fields exact; status Draft (not submitted).
Proposed partial credit
45% correct sailing; 35% cargo fields; 20% not submitted.
R05.09.01.03-T20 Build, repair and tender a weekly transport plan in OTM
Original instruction
In Oracle OTM (training), bulk-plan the 40 order releases in input/week42_orders.xlsx. The plan leaves 3 constraint failures (overweight shipment, missed appointment, hazmat mix). Fix them by moving order releases between shipments in the shipment workbench, re-rate, tender all shipments to the selected carriers and publish the plan report to output/plan.pdf.
- Platforms
web· Any OS (web)- Applications
- Oracle OTM · Excel
- Interfaces
brow· Browser- GUI need · task setting
G2· GUI-essential- GUI need · ordinary access
G2· GUI-essential- Interaction primitives
drag· Drag & dropclick· Point & clicknav· Pan/zoom/scrollmulti· Cross-app- Work product
WP08· App/system state- Conditions
long· Long-horizon- Proposed difficulty
- 5 / 5
Proposed verifier
OTM DB: all 40 releases on shipments; no weight, appointment or hazmat-compatibility violations; total plan cost <= reference +5%; all shipments tendered; report PDF present. Workbench drag events required (integration user disabled).
Proposed partial credit
20% bulk plan; 40% three failures fixed; 20% cost within 1.05x reference; 20% tendered and published.
Mapped research references
No benchmark reference is assigned to this class in the current mapping.