AI Yard Management System: How It Tracks Containers and What It Costs
A yard management system (YMS) is software that tracks every container, trailer and truck between the gate and the dock or stack. It records arrivals, assigns slots, issues move orders and measures dwell time. The yard team works from a live record instead of paper and memory.
An AI yard management system adds computer vision and tracking to that record, so containers are identified and followed without manual typing.
When Savvycom started work for a global logistics provider in March 2024, its container yards had none of that. Each yard handled 400 to 500 container movements a day, with more than 500 containers on the ground at once. Staff relied on manual records to find containers. Moves inside the yard were not recorded, so a container that was moved could drop out of the record entirely.
1. How is a yard management system different from a WMS, TMS or TOS?
A yard management system covers the space outside the warehouse: the gate, the yard and the moves to docks or stacks. A warehouse management system (WMS) manages stock inside the building. A transportation management system (TMS) plans transport between sites, and a terminal operating system (TOS) runs a port terminal.
Buyers mix these up because the systems share data. A truck booked in the TMS arrives at the yard gate, its container waits in the yard, and the cargo is received in the WMS. Each system owns a different part of that journey.
| System | What it covers | Main users | What it records |
|---|---|---|---|
| Yard management system (YMS) | Gate to dock or stack, inside one yard | Gate clerks, yard controllers, yard truck drivers | Location, status and dwell time of each container, trailer and truck |
| Warehouse management system (WMS) | Stock inside a warehouse | Warehouse and inventory staff | Stock by item and bin, picking, put-away |
| Transportation management system (TMS) | Transport between sites | Planners, carriers | Loads, routes, carrier bookings, freight cost |
| Terminal operating system (TOS) | A port or rail terminal: quay, stack and gate | Terminal operators | Vessel plans, stack positions, crane and gate moves |
Dwell time, the number every YMS reports, is how long a container or trailer stays in the yard from gate-in to gate-out. It is the first sign of congestion and of charges from shipping lines.
Many container depots and logistics yards run without a TOS, and a WMS does not see past the dock door. The client in the case below had no specialized yard system at all. Our guide to logistics software development covers warehouse and transport systems and how they connect to the ERP. The transportation management software guide goes deeper on the TMS side.
2. How does a yard management system work, and which features matter?
A yard management system works as a sequence from gate-in to gate-out. It identifies each truck and container, assigns a slot and issues a move order. It then confirms where the container went and keeps the record live until exit. Steps that rely on someone typing a number are where the record drifts.
Gate check-in and move confirmation carry the most risk, because both depend on a person entering the right number at the right moment. As a simple illustration, a 1% error rate on 400 moves a day is four wrong records every day. Each one sends someone searching later.
Features follow from that sequence. Truck appointments spread arrivals across the day, and check-in speed decides the queue outside the gate. A live yard map should be searchable by container number, booking or customer. Move orders need confirmation, not assumption. Dwell time alerts flag ageing containers before detention charges start. Warehouse yards with many doors also need dock scheduling; container depots care more about stacking rules and identification accuracy. Integration with the WMS, TMS or ERP, plus an audit trail of who moved what, closes the list.
3. Where does AI change a yard management system?
AI changes yard management at the two steps that depend on manual entry: identifying containers at the gate and confirming moves inside the yard. Computer vision reads container IDs and truck plates from camera images, and object tracking follows each container, so the record updates without anyone typing.
- Reading IDs at the gate. Optical character recognition (OCR) reads the container number, trailer ID and truck plate from the camera image. In the Savvycom project, this replaced manual data entry at entry and exit.
- Following moves inside the yard. A detection model finds containers in each frame, and a tracker keeps the same identity as they move. Internal moves then appear on a live 2D yard map instead of going unrecorded.
- Checking placement against the move order. If a container is not where the order says, the mismatch shows up during the move, not days later during a search.
- Sending doubtful reads to a person. Every AI read carries a confidence score. Savvycom’s Operational AI in Logistics whitepaper sends predictions below 85% confidence to human review. They are not written straight into the record.
AI cameras are not the only way to track a yard. The right choice depends on who owns the assets and how often they move.
| Tracking method | How it identifies assets | Works well when | Watch out for |
|---|---|---|---|
| Manual records and handhelds | Clerks type or scan IDs | Low volume, few internal moves | Typing errors; internal moves go unrecorded |
| Barcode or RFID tags | Readers at the gate or on equipment | You own the trailers and can tag them | Containers from many shipping lines rarely carry your tags |
| Real-time location system (RTLS) on yard equipment | GPS or ultra-wideband locates the yard truck or stacker | You need precise equipment positions | Tracks the machine, then infers the container |
| AI cameras (computer vision and OCR) | Reads container IDs and plates from images | Mixed container owners and high move counts | Needs camera coverage, lighting and model tuning per site |
Before you commit, ask any vendor to show recognition working on your own camera footage. Test it at night, in rain and on every plate format you receive. Most Thai plates, for example, use Thai script, so a model trained only on Latin characters will miss them.
Ports are moving in the same direction. The World Bank Container Port Performance Index for 2025 covers 403 container ports. It points to digital platforms linked to customs and logistics partners as one way to improve competitiveness and resilience.
4. How much does a yard management system cost?
Published prices range widely. Vantage YMS starts at $149 a month for phone-based gate check-in. Peripass lists a QuickStart plan from EUR 1,299 per site per month, billed annually. Most enterprise yard vendors quote on request. A custom AI yard system is priced by team size, project length and hardware instead.
Both prices come from the vendors’ own pricing pages, checked in November 2026. The Vantage entry plan covers 2 yards and 2,000 gate transactions a month. The Peripass plan includes 75 trucks a day. Few yard management vendors publish prices, so expect a quote rather than a price list from the larger platforms. The cost also changes shape depending on whether you buy, extend or build.
| Cost item | Buy (SaaS YMS) | Extend or build (custom AI) |
|---|---|---|
| Software | Subscription per site per month | Engineering team for design, build and testing |
| Hardware | Gate kiosks, handhelds, sometimes cameras | IP cameras at the gate and across the yard, plus cloud or edge compute for inference |
| Integration | Standard connectors; custom ones often cost extra | APIs to your WMS, TMS or ERP, built into the scope |
| Running cost | Included in the subscription | Cloud hosting, monitoring and model retraining |
| Ownership | The vendor owns the product | You own the code, models and data |
For a custom build, engineering time is the main cost line. Savvycom’s 2026 rates for Vietnam-based teams are $22 to $48 per hour for most roles, with senior AI and data roles reaching $55. At about 160 hours a month, one engineer-month comes to roughly $3,500 to $7,700. Your budget is that figure multiplied by team size and months, plus cameras and cloud.
Scope five things early, because they drive most of the variation. First come the number of yards, gates and camera positions, and the lighting at each site. Then come the systems that must receive yard data, the accuracy target you set, and the plate formats across the countries you serve.
5. Should you buy, extend or build a yard management system?
Buy a SaaS yard management system when a product already fits your gate, yard and dock process. Extend a platform with custom AI when the product works but identification at the gate or inside the yard does not. Build when you run high-volume container yards with mixed container owners and in-house systems no product connects to.
| Approach | Fits | Trade-off |
|---|---|---|
| Buy: SaaS YMS | Warehouse and trailer yards with standard flows, teams starting out | Fast and lower upfront cost, but limited customization and dependence on the vendor |
| Extend: platform plus custom AI | A YMS or WMS is in place but gate and move records are still unreliable | Needs moderate technical skill and integration effort |
| Build: custom AI yard system | Container yards with mixed owners, many moves and in-house systems | Higher cost, a longer timeline and ongoing model maintenance |
Extending or building is probably not worth it when:
- The yard sees a few dozen trucks a day and flows are steady.
- A SaaS YMS already covers your process.
- There is no budget or network for cameras across the yard.
- Yard procedures differ across teams or sites and are not written down yet.
The last point catches many projects. The whitepaper lists three readiness factors: mature data infrastructure, standardized processes, and teams able to act on AI output. If procedures vary across teams or sites, a model learns the variation, not the rule.
6. What results can an AI yard management system deliver, and how long does it take?
Savvycom built an AI yard management system for a global logistics provider whose yards handle 400 to 500 container moves a day. The nine-month project ran from March to December 2024. It reached 95% accuracy in container placement and movement tracking and cut container search and retrieval time by 60%.
Savvycom delivered AI camera recognition at entry and exit, real-time tracking across the yard, a 2D digital yard map, AI-validated move orders and operational dashboards. Vision runs on YOLOv8 for detection, PaddleOCR for reading IDs and plates, and DeepSORT for tracking. Python and .NET services sit on PostgreSQL, MinIO, Kafka and Prometheus, with REST APIs on Microsoft Azure.
Results were measured against the client’s earlier manual process. Read the full AI yard management case study.
The system targets the three points where manual yard records drift: identity at the gate, moves inside the yard and placement checks. These numbers describe this client’s yards, volumes and camera setup. Treat them as a reference point, not a forecast.
If you plan a similar project, measure these before you start, so the result can be proven later:
- Average time to find a container, timed on a sample of searches.
- Moves per day, split into gate moves and internal moves.
- Misplaced or lost containers per week.
- Dwell time by container type and customer.
- Truck queue time at the gate at peak hours.
A safe rollout runs the AI alongside the existing records first and compares both before switching. The whitepaper calls this a parallel run, or shadow mode when the AI only observes without changing operations. Start with one gate, then extend camera coverage into the yard once read accuracy holds on your own footage.
Frequently asked questions
What is the difference between detention and demurrage?
Demurrage is charged when a container stays inside the port terminal beyond its free time. Detention is charged when a shipper or consignee keeps a carrier’s container outside the terminal beyond its free time. A yard management system with dwell time alerts helps you move or return containers in your own yard before detention starts.
What is a gate OCR system?
A gate OCR system uses cameras at the yard or terminal gate to read container numbers, truck plates and sometimes seal numbers as trucks pass. It replaces manual typing at check-in and check-out. Readings with low confidence go to a gate clerk to confirm, so errors are caught before they enter the yard record.
Do you need a yard management system if you already have a TOS?
Usually not inside the terminal itself, because a terminal operating system already manages the stack, quay and gate. A separate yard management system fits container depots, inland yards and warehouse yards that run without a TOS. It also suits operators who need one view across several yards.
How long does it take to implement an AI yard management system?
An AI yard management system takes months rather than weeks, because cameras, models and integrations must be tuned to each site. Savvycom’s project for a multi-yard logistics provider took nine months, covering camera recognition, tracking, a digital yard map and dashboards. More sites and integrations lengthen the timeline.
Planning a yard, warehouse or transport project in Asia-Pacific? Talk to Savvycom’s logistics software team about your yard volumes and existing systems.
Looking for a Trusted Tech Partner That Delivers Business Impact?
Savvycom is a Vietnam-based software company with 16+ years of delivery, 100+ enterprise clients and work in 30+ countries. Our smart logistics team works on yard, warehouse, transport and AI systems for operators across Asia-Pacific.
Explore: Smart Logistics Solutions
