- Home page
- h::AI – AI for manufacturing a...
h::AI® – AI for manufacturing and logistics
AI on the shop floor. Ask your system about your data.
h::AI is an artificial intelligence module integrated with HDF MES and HDF WMS. Ask about production, downtime, scrap or stock levels and get an answer based on data from your system. No digging through reports.
Building software for industry since 1999 · Demo on your own data
Google reviews(4.7)Built into MES and WMS
Works inside the system your team already uses.AI on-premises or in the cloud
h::AI runs on your own server or uses external AI services.You decide who has access
You give h::AI access to selected people, e.g. planners and managers.AI advises, people decide
h::AI makes suggestions and flags problems; decisions stay with you.
Introduction
What is h::AI?
h::AI® is HDF's brand of artificial intelligence solutions. It covers modules integrated with our industrial software as well as AI solutions tailored to individual company needs.
In HDF MES and HDF WMS, h::AI lets users quickly get production and warehouse information, using the data already available in the system.
Reports
Why h::AI if you already have reports?
Reports show summaries prepared in advance. h::AI lets you ask about a specific problem and get an answer based on the available data – without hunting for the right report or setting filters.
| Situation | Without h::AI | With h::AI |
|---|---|---|
| A question no report answers e.g. “How much scrap did the night shift produce on product X over the last two weeks?” | Someone has to prepare a summary; the answer comes tomorrow | You ask and get the answer right away |
| A quick check before a briefing or decision | Switching views, setting filters, exporting | One question – no filters, no clicking through views |
| Requests for “quick” summaries | Take up planners' and IT's time | Managers check for themselves; the team gets on with its work |
| Data from several views at once e.g. production, scrap and downtime in one summary – in the extended configuration | Manually combining data from different screens, or ordering an extra report or module | h::AI pulls data from different views into one answer, with no extra modules to order |
Reports and dashboards stay. h::AI fills the gaps where a ready-made report is not enough.
Planning
Planning people and machines in HDF MES
A machine breakdown, staff absence or a rush order can call for a quick change of plan. In the extended configuration, h::AI helps find possible solutions and assess their impact on production.
Machine breakdown
h::AI helps rework the plan after a machine breakdown and shows how to limit delays.Rush order
When a rush order comes in, h::AI helps check which orders can be moved and how that will affect the plan.Staff absences
When you are short-staffed, h::AI helps adjust crewing and the schedule to the resources available.“What-if” simulation
Check the impact before you decide: what happens to the plan if machine X is down until tomorrow, or a shift is three people short?
For planners
Less searching, more time for decisions
Planners need fast access to information about orders, machines and production progress. With h::AI, they can ask questions without going through multiple screens and reports.
In the extended configuration, h::AI also helps analyze schedule changes and assess their impact on production.
h::AI does not replace the planner. It takes over the tedious data hunting, leaving more time for what AI cannot do: knowing the people, the machines and the company's priorities.
Principle: h::AI proposes a plan change; a person makes and approves the decision.
Planning and simulations are available in the extended configuration.
Modules
Where h::AI works
h::AIh::AI in HDF MES
Ask about production the way you would ask a process engineer: how much was produced, where the scrap is, which machine was down.
- “How is production going today?”
- “Which machine had the most downtime this week?”
- “How much scrap was there on the night shift?”
h::AIh::AI in HDF WMS
Stock levels, locations, picking. Ask the system instead of scrolling through lists.
- “How much of product X do we have, and where is it?”
- “Which orders are waiting to be picked today?”
- “What has been sitting in the warehouse for more than 90 days?”
h::AI also supports load planning in HDF ATL, a system in development since 2018: it flags errors in the plan and suggests corrections. See HDF ATL →
How it works
Working with h::AI
You don't need to know the report layout or complex filters. Pick a topic and ask your question in plain language.
Pick a topic
Choose the area you're interested in, e.g. machines, production or warehouse.Ask in your own words
Ask in plain language, the way you would ask a colleague.h::AI answers or asks back
You get an answer based on system data. If needed, h::AI asks you to clarify the question.
- Standard configuration
Conversation and tables
Answers as conversation and tables, in every configuration. - Extended configuration
Charts and summaries
In the extended configuration, h::AI can analyze data from different areas of the system and present results as charts, tables and summaries.
We match the scope to your needs and infrastructure during the pre-implementation analysis.
Comparison
Why isn't a regular AI chat enough?
Many companies try AI by pasting system data into a public chat. That works as a start, but in day-to-day production its limits show quickly.
| AI chat with pasted data | h::AI in HDF MES and WMS | |
|---|---|---|
| Data preparation | Someone has to export the data and paste it in | You pick a topic, the system prepares the data |
| Freshness | Data as of the export | Data straight from the system, refreshed according to your MES/WMS configuration |
| Where data goes | To an external AI service | On-premises or to a selected AI service — depending on configuration |
| Who has access | Anyone who has the export | Access follows the permissions set in the system |
| Unclear question | Answers even when it didn't understand | Can clarify the question using the system context |
| Missing information | Works only with what you paste | Selects the date range and data it needs from the system on its own* |
| Cost | Grows with the number of questions and users | No per-question fees in the on-premises option |
* in the extended configuration
Implementation
How we implement h::AI
We implement h::AI together with HDF MES or HDF WMS, or add it to a system you already run.
Pre-implementation analysis
We get to know your processes and data. We identify the areas where AI will bring real benefit and match the configuration to your needs and server capacity.Solution design
We define the conversation topics, the scope of data and the people who will use h::AI.Launch and testing
We launch h::AI on your data. Selected users check the answers before the module goes into daily use.Rollout and tuning
We train users. In the initial period, we tune the answers to your terminology and questions. The scope of tuning and further development depends on the chosen configuration.
Demo
See h::AI on your company's data
We'll show you h::AI using examples from your own data. And we'll tell you straight whether AI makes sense for you.
[email protected]Custom solutions
Custom AI assistants and agents
Not every process in a company starts and ends in MES or WMS. We design custom AI assistants, agents and chatbots: solutions that take over repetitive work with documents, inquiries and data from different systems.
Every project starts with checking whether AI makes sense for the process. If it doesn't, we'll tell you straight.
Example applications:
- company AI assistant: a chat for your team that knows your documentation, procedures and data, and works within your security rules
- AI chatbot for your website that answers customers based on your site content and offer
- first-line handling of tickets and customer questions
- preparing quotes from inquiries
- finding information in company documentation and data
We deliver website chatbots as projects tailored to your offer, not as off-the-shelf template subscriptions. The chatbot answers only from the content you specify. When it doesn't know the answer, it refers the user to your team instead of guessing. It also tells users they are talking to AI, as required by the EU AI Act.
We integrate with SAP, Oracle, Microsoft SQL Server, MariaDB and MySQL, among others, as well as with data in Excel, PDF and CSV files. We work with both locally hosted models and cloud services, including OpenAI (GPT), Anthropic (Claude), Google (Gemini) and Mistral AI.
Describe your process
Frequently asked questions about h::AI
How is h::AI different from a regular AI chat?
A regular AI chat does not know your system. To ask it about production, you first have to export the data and paste it in. h::AI is built into HDF MES and WMS: you choose a topic, ask a question, and the answer is based on data from your system, with user permissions respected.
Why do I need h::AI if my MES already has reports?
Reports answer the questions that were anticipated when they were designed. h::AI answers the questions no ready-made report covers, right away and without asking anyone to prepare a summary. Your reports stay; h::AI complements them.
Does AI in manufacturing require sending data to the cloud?
Not necessarily. h::AI can run on a server in your company or use a selected cloud AI service. We choose the option that fits your company's security policy and budget.
Does h::AI make decisions instead of people?
No. h::AI analyzes data, flags problems and suggests solutions, but decisions are always approved by a person. The same applies to planning: h::AI proposes changes to the plan, but the planner puts them in place.
Who in the company uses h::AI?
You decide who gets access to the module. In HDF MES, it is most often used by production planners, as well as managers and senior management. Access follows the permissions set in the system.
How much does it cost to implement AI in manufacturing?
h::AI is implemented in your system, so we quote each deployment individually. The cost covers the h::AI module, infrastructure – your own server or a cloud AI service – implementation with tuning to your data, and ongoing support. We agree on the scope during the pre-implementation analysis so that it fits your budget.
Does h::AI work from day one?
Yes, but it reaches full answer quality after a tuning period. Every company has its own terminology, processes and questions, so after implementation we adapt h::AI to the way your team works.
Can I add h::AI to an MES or WMS from another vendor?
Not as a ready-made module, because h::AI is built into HDF MES and HDF WMS. For systems from other vendors or in-house solutions, we build a custom AI assistant based on data from your system. The scope depends on how the system exposes its data, and we agree on it during the pre-implementation analysis.
Can we have our own AI chat in the company?
Yes. We build company AI assistants that work with your documents and data. They can run locally on your server or use cloud services under business terms. During the analysis, we help you choose the option that fits your security requirements and budget.
Do you build custom AI agents and chatbots?
Yes. Besides h::AI in HDF MES and WMS, we design AI assistants, agents and chatbots, for example for websites, tailored to your company's processes. More about custom solutions.