HOW ORBIT WORKS
AI agents that never stop looking. Meteorologists who never stop backing your judgement.
HOW ORBIT WORKS
Maritime expertise and your company's rules, working on every voyage.
Most voyage software stores data and waits for you to ask the right question. Orbit is built the other way round. It works through maritime data, tools and encoded expertise, follows your objectives and rules, and brings you the decisions worth making. This page shows what happens underneath.
The four layers
Four layers, working to your rules
Orbit has four layers. Data at the bottom. Tools that calculate on it. Skills that encode how maritime experts approach a problem. And people at the top: the meteorologists and master mariners of Orbit Marine Services.
Orbit’s AI works across the lower three. It checks the data, runs the tools and follows the skills, coordinating them toward the goals you have set. Everything it does is shaped by your company’s context: your objectives, your definitions, your operating rules.
You set the scope and make the call. Orbit does the work.
Data layer
Years of integrations, meshed into one picture
Tools layer
Monitoring on your terms
Orbit monitors what you ask it to monitor, within the rules and thresholds you choose. Nothing more, nothing less. Inside that scope it works continuously, rerunning the simulation when the forecast changes and comparing performance with the warranty. It raises something when your situation calls for it, and otherwise stays out of your way.
You ask Orbit to keep a voyage on schedule at the lowest fuel cost. A low-pressure system deepens on the planned route three days ahead. Orbit reruns the routing and finds a northern deviation: 40 nm longer, clear of 7 m waves. Both options appear side by side with fuel, ETA and risk.
Skills layer
Maritime expertise, written down
A skill is encoded maritime procedure: the steps an experienced operator follows for a specific task. Each skill knows which tools to call and which data it needs along the way. General AI models can read a contract. Orbit’s skills know which clauses drive the voyage result, what to calculate and what to do about it, and they apply your company’s definitions as they go.
The skill reads the laycan and notice of readiness terms in the charter party. Say NOR can only be tendered until 16:00 on a Friday, and the vessel is heading for an 18:30 arrival. Laytime may not start until Monday: two days that would otherwise count. Orbit flags it five days out, calculates the speed needed to arrive before 16:00, and sets the extra fuel against the five days.
The skill reads the speed and consumption warranty, including its good-weather definition. Each day it identifies which periods qualify, compares actual speed and consumption with the warranty, and shows whether there is a potential underperformance claim. Owners see their exposure early. Charterers see what they can claim.
Laycan and NOR tendering
The skill reads the laycan and notice of readiness terms in the charter party. Say NOR can only be tendered until 16:00 on a Friday, and the vessel is heading for an 18:30 arrival. Laytime may not start until Monday: two days that would otherwise count. Orbit flags it two days out, calculates the speed needed to arrive before 16:00, and sets the extra fuel against the two days.
CP performance.
The skill reads the speed and consumption warranty, including its good-weather definition. Each day it identifies which periods qualify, compares actual speed and consumption with the warranty, and shows whether there is an underperformance claim and how large. Owners see their exposure early. Charterers see what they can claim.
Human layer
Meteorologists and master mariners, on call
Some decisions need someone who has seen the situation before. Orbit Marine Services has meteorologists on call round the clock from Norrköping, Sweden, and master mariners on the team.
Orbit can bring them in when a voyage needs human judgement, for example a routing decision through a developing storm. You decide how: let Orbit contact Orbit Marine Services directly when a voyage crosses a threshold you have set, or have it recommend a review and leave the decision to you.
Orbit does the monitoring and the first analysis. The experts bring the judgement.
Human layer
Expert review, on your terms
Skills layers
Tools layer
Data layer
Applies
Monitors
Checks
Your objectives, definitions and rules
Your company, built in
Your company's rules, alongside maritime expertise
Generic AI answers the way it would answer anyone. Orbit answers the way your company works. It applies your objectives, definitions and operating rules alongside its maritime expertise, in every check, calculation and recommendation.
01.
Objectives — What matters most
Arriving on time, saving fuel and cost, staying on a CII trajectory. Orbit weighs its options against the objectives you set for each voyage.
02.
Operating rules — How you work
Which decisions need approval, which savings are worth an alert, when Orbit Marine Services may be involved. Orbit stays inside them.
Two companies with the same vessel on the same voyage can get different advice from Orbit, because their rules differ.
One voyage, four decisions
What Orbit brings you on one voyage
A Supramax, laden, on an 18-day voyage. Each event is Orbit at work, using different skills, tools and data to reach the goals you have set. Illustrative example.
Day 3
Day 7
Day 11
Day 16
Built differently
Built differently from the software you know
| Traditional voyage software | Orbit | |
|---|---|---|
| Who starts the work | You open a module and run the analysis | Orbit works continuously, within the scope you set |
| Data quality | Errors surface when a number looks wrong | Reports are checked on arrival, fixes proposed |
| Maritime knowledge | Lives in your team's heads and spreadsheets | Encoded as skills, applied with your company's rules |
| What you get | Charts and tables to interpret | A proposed action with its estimated cost or saving |
| Views and reports | Fixed modules, the same for every user | Built for your role, and new ones on request |
| Expert help | A separate service, by email or phone | On call; Orbit brings them in on your terms |
| AI models | Fixed at purchase, or none at all | Evaluated and upgraded by OrbitMI as models improve |
A glass box
You set the rules. You see the reasoning.
Orbit works inside limits you define, and every step it takes is open to inspection. We call it a glass box.
01.
You set the scope
What Orbit should monitor, how large a saving is worth an alert, and when it may involve Orbit Marine Services on its own. Orbit works to your settings, and you can change them at any time.
02.
You see the trail
Every recommendation shows the data it used, the tool that ran and the skill it applied, with links back to the source. Check it, rerun it with different inputs, or overrule it.
03.
You make the call
Orbit checks, analyses and flags on its own. Nothing changes your voyage without your approval.
Keeping pace with AI
Better AI, without changing anything on your side
AI models improve fast, and the best model for a maritime task today may not be the best in a year. Keeping up is OrbitMI’s job.
We test
New models are evaluated against maritime tasks: reading charter party clauses, interpreting noon reports, reasoning about weather and routing.
We upgrade
When a model does the job better, OrbitMI moves Orbit onto it and checks the results before you see them.
You keep working
Your skills, rules and workflows stay as they are. You get the improvement without switching providers or rebuilding anything.
FAQs
AI power. We handle the complexity.
You focus on maritime operations.
How does OrbitMI use AI in the maritime industry?
OrbitMI uses AI in the maritime industry to monitor voyages, clean data, and to assist the user with data analysis and reporting – not to replace people. AI agents watch the data your fleet already produces – noon reports, AIS, weather and your own systems – flag what has changed and propose specific adjustments or actions. The agents can look for specific issues or opportunities or help you with your analysis. Human meteorologists can review the recommendations, so expert advice is on hand before you act.
Does OrbitMI use AI or human experts?
Both. AI agents monitor your voyages continuously, flag alerts and propose specific adjustments or actions. Behind the AI is an expanding team of human meteorologists operating from Sweden with round-the-clock support, and OrbitMI’s Marine Services team of meteorologists and master mariners can review routing advice before your crew act on it.
Where are OrbitMI’s meteorologists based?
OrbitMI’s meteorologists operate from Sweden, including its Norrköping office, and provide round-the-clock support. They review routing recommendations, monitor each voyage daily, and keep in direct dialogue with the Master when a recommended strategy needs discussing.
Does OrbitMI replace my existing systems, or connect to them?
Both. It can replace your voyage optimization and compliance monitoring systems, but it also connects to the systems you want to keep. OrbitMI links to the systems you already trust and adds what you’re missing – live weather routing, compliance tracking and fleet performance – as one layer that fills the gaps without replacing what already works. Integrations run through open APIs and standardized data exchange, scoped around your workflows, data access and subscriptions.
Why not just use Claude or ChatGPT?
We have a strong data team. Why not build it ourselves?
What does OrbitMI add?
Your data and maritime data in one place, checked and ready to use. Built-in maritime methods for CP performance, weather routing and report quality, plus many more. An agent puts them to work for your team every day.
Can the numbers support a charter party claim?
What about AI errors?
Our reports vary in quality. Can you handle that?
What does the agent do?
Will everyone see the same numbers?
Yes. Commercial, operations, technical and management work from the same workspaces and the same methods. Answers don’t depend on one person’s prompts.
Who supports us?
Mariners and meteorologists built our methods and check them. They help your team when a result needs a second look.
What if a better AI comes out?
New models come out every few months. We test them on maritime tasks and switch when one is better. Nothing changes for you.
Is our data safe?
Your data stays in your account. Access is by role, and your data is not used to train AI models.
Will AI costs grow?
Not like using AI directly. In OrbitMI, saved workspaces refresh without AI, and only new questions use AI.