A-101 · Swift Haulage
51 min26 pallets · palletised
Warehouse & loading-bay planning
DockThimble helps warehouses schedule vehicle arrivals around realistic loading and unloading times. Machine learning predicts appointment duration from shipment size, handling needs, staffing and past turnarounds — for dock planners to confirm.
Dock board · 16 vehicles · 6 doors · 3 teams
fictional day
The problem
Most warehouses book every vehicle into the same slot length. A palletised trailer is out in under an hour; a floor-loaded container takes nearly ninety minutes. The slots look tidy on the booking sheet and fall apart in the yard.
37% of past turnarounds at this fictional site ran over the one-hour slot.
Average minutes at the door by load type. The dashed line is the one-hour booked slot.
Duration predictions
DockThimble learns from your own turnaround records: pallet count, how the load is packed, chilled handling, document checks and crew size. Every estimate carries the model’s typical error on visits it never saw — ±3.9 minutes here.
A-101 · Swift Haulage
51 min26 pallets · palletised
A-102 · Coastline Logistics
106 min22 pallets · floor-loaded · document check
A-105 · Prime Movers
117 min30 pallets · floor-loaded
Appointment planning & congestion warnings
Each vehicle is matched to a compatible door and a free team. DockThimble then simulates the day and flags half-hours where two or more vehicles will be waiting — so bookings can move before the yard fills up.
Booked in one-hour slots
Planned with predicted durations
Vehicles waiting per half-hour on the same fictional day. Red marks a congestion warning.
Schedule adjustments
The supervisor records a late arrival; DockThimble re-plans doors and teams around it and suggests which bookings to move. The planner decides what to agree with carriers.
Kola Transport (A-106) · running 75 min late
Turnaround reporting
| Visits | Avg minutes | Over slot | |
|---|---|---|---|
| Northbridge Freight | 51 | 64 | 45% |
| Coastline Logistics | 56 | 61 | 38% |
| Prime Movers | 43 | 60 | 40% |
| Swift Haulage | 63 | 59 | 37% |
| Kola Transport | 46 | 58 | 28% |
| Delta Reefer | 61 | 57 | 33% |
| Visits | Avg minutes | Over slot | |
|---|---|---|---|
| Document check | 56 | 78 | 66% |
| Chilled | 59 | 59 | 34% |
| Standard | 205 | 55 | 29% |
From 320 fictional past turnarounds.
How it works
Upload past turnaround records — arrival, start, finish, pallets, load type.
Add tomorrow’s bookings by hand or from your booking sheet.
Durations are predicted; doors and teams are matched; queues are flagged.
Accept rebooking suggestions, record late arrivals and publish the plan to the dock team.
Suggestions are for the planner to accept; the published plan is theirs.
Durations learn from your own dock, not an industry average.
Every estimate comes with its typical error.
DockThimble never contacts carriers on its own.
Why it matters
Share of stops in 2023 at which US truck drivers were detained — held beyond two hours waiting for loading or unloading.
ATRI, Costs and Consequences of Truck Driver Detention (2024)Productive hours the US for-hire trucking industry lost to detention in 2023.
ATRI (2024)Revenue the US for-hire trucking industry lost to detention at shipper and receiver facilities in 2023.
ATRI (2024)Who it’s for
Many clients, many carriers, one set of doors.
High daily volumes where a bad morning ripples all day.
Mixed loads and floor-loaded containers that never fit a standard slot.
Inbound materials and outbound orders competing for the same bays.
Pricing
Illustrative pricing for a proposed product, to be agreed with pilot customers.
$290per warehouse / month
$240per warehouse / month, 3+ sites
Custom
Technology
DockThimble’s work changes by the minute: appointments are booked, vehicles run late and doors free up early, while turnaround records and delivery documents pile up behind every visit. Designing on AWS lets record storage, document reading, duration models and scheduling workflows run as managed services, and keeps each warehouse’s data isolated as customers grow from one site to a network.
Amazon S3
Turnaround histories, booking exports and scanned delivery documents land in per-warehouse storage.
Amazon Textract
Delivery notes and CMRs are read for pallet counts, load type and handling requirements, so appointments carry real details.
Amazon SageMaker AI
A duration model trained on each site’s turnarounds estimates loading and unloading time, with its error on held-out records.
Amazon DynamoDB
Appointments, door types, teams, arrivals and planner decisions are stored per warehouse and day.
AWS Lambda
Lambda re-plans doors and teams when a booking changes or a vehicle is late, and raises congestion warnings.
Across every stage
Proposed, not deployed. This website and the dock planner run in your browser on fictional data.
Team
Founder
Samuel founded DockThimble after watching yards fill with trucks booked into one-hour slots for jobs that take two — and dock teams standing idle an hour later. Samuel leads product, customers and every pilot.
Progress & pilot programme
DockThimble is in its building stage and opening its first pilots. A pilot covers one warehouse: you upload past turnaround records, your team enters appointments, and we compare predicted and actual times every day.
By Samuel Akinwale, to make dock appointments match how long loading really takes.
Duration model from your own turnarounds, door and team matching, queue warnings, planner approval.
Predictions, door plan, congestion, late-arrival suggestions and reports shown on a fictional warehouse.
Records intake, document reading, duration models, appointments and scheduling workflows designed.
One warehouse per pilot: appointments entered by your team and uploaded turnaround history.
FAQ
A few months of past turnarounds — arrival, start and finish times, pallets and load type — plus your doors and team sizes.
No. The first version takes appointments entered by your team. Integrations come later, on request.
No. It suggests rebookings; your team agrees them with carriers.
Not yet. It is a proposed product; the dock planner is a prototype running in your browser on fictional data.
Contact
Questions about the product, the pilot programme, data imports or how warehouse records are handled. Messages go directly to the founder.