Today
Many of your machines at customer sites record operating data: operating hours, loads, temperatures, current draw. Some of it arrives by remote access; the rest is read out at the annual inspection and ends up as a file in the work order. Mostly nobody looks at the data until the customer calls because something has stopped, and then it becomes an emergency call-out instead of a planned visit.
An example
Conveyra Fördertechnik looks after about 1,400 cranes and hoists at customer sites; 300 send their operating data daily, and for the rest the technician reads out the data logger at the annual inspection. Monday's list shows three cases. On a 10 t overhead crane at a steel fabricator, current draw and temperature of the hoist motor have been rising for three weeks at the same load, just as before two earlier failures caused by dragging brakes on the same series. A hoist in a foundry has used 91 % of its theoretical service life, and a hall crane has logged repeated overload cut-outs on the early shift for a week. The dispatcher books the brake replacement with the part for Thursday, the general overhaul quote goes out after approval, and the customer with the overloads gets a call.
Your data, wherever it lives
Today operating data arrives by remote access in a database, read-outs from the annual inspection sit as files in the work order folder, service history in the CMMS, contracts in the CRM and spare parts in the ERP. demi reads these sources, understands which values belong to which serial number, and links them into one curve per machine, with your rights. AI spots trends across the whole installed base in minutes, and your service team decides who gets a call.
Where the data lives today
- Machines & sensors
- Databases
- Folders & drives
- Tickets
- CRM
- ERP
How demi solves it
Three layers in this order: connect the data first, then the rights, then the AI.
- 1 · ConnectData Layer
The case needs data from machines and data loggers, a time-series database, the CMMS, CRM and ERP. demi replaces nothing and brings together, per machine at each customer, what arrives by remote access, what was read out at inspections and what sits in service history, contract and spare parts stock. Every serial number gets one continuous curve, whichever way the values came in.
- 2 · PermitPermission Layer
Each customer's data is visible only to the responsible service team, and every customer sees only their own machines in the portal. demi uses patterns across several customers' machines only where your contracts with them allow it. Service approves every message to a customer before it goes out.
- 3 · ActIntelligence Layer
The agent learns each machine's normal state at its load and shift pattern and looks for trends over weeks, not single outliers. It compares unusual curves with what the service history showed before earlier damage and proposes visit, spare part and date, with the curve and comparable cases as evidence. It never touches the controls; it only reads.
How it runs
- 1
demi collects the condition data of your machines at customer sites, daily by remote access or as a file from the last inspection.
- 2
Every machine gets one continuous curve, matched with serial number, year of build, contract and service history.
- 3
If vibration, temperature or current draw drift over weeks, the agent compares the curve with earlier damage.
- 4
Every Monday a list is ready: which machine, which risk, which visit with which part, with curve and evidence.
- 5
Service calls the customer or sends the approved message, and the visit is planned before the machine stops.
What changes
You call the customer before their machine stops
Planned visits instead of night-time call-outs
The right spare part travels on the first visit
What does it bring you?
Check it with your own numbers
Every starting value is an assumption you can change. The calculation sits right below, step by step.
Your numbers
Result
+144 hfreed per year
Assumptions and calculation
demi reviews every machine's data and flags only what stands out; service decides on the visit.
20 min → 14 min per case
How we calculate
120 machine data reviews × 20 min × 30% × 1.0 (manual) × 12 months = 8,640 min = 144 h
Value per year, before demi costs
Freed per year+144 h
Assumption: labour cost of €49.50 per hour in German manufacturing in 2025 (Destatis), plus workplace, IT and overheads.
= Value per year, before demi costs
+€9,360
gained per year
144 h ÷ 1,760 h per role = 0.1 full-time roles
An estimate, not a promise. The starting values are assumptions for an industrial company with 150 to 500 people.
Combine with more cases and get the packageQuestions about this use case
Only some of our machines at customers are connected. Is that enough?
Yes. Where there is no remote access, demi reads the files read out from the data logger at the annual inspection or during a visit. Trends then show up less often, but they do show up. The list also tells you which machines have sent no values for a long time.
Does demi interfere with the controls at the customer's site?
No. demi only reads condition data, writes nothing back and acknowledges no messages on the machine. How the data gets to demi, through your remote access, an export or files from the work order, we clarify in a conversation.
How does demi track the remaining service life of hoists?
From operating hours and load spectrum, as FEM 9.755 describes for serial hoists, or straight from the value the data logger already calculates. demi shows when the theoretical service life will be reached and reminds you in time to quote the general overhaul. The assessment at the inspection stays with the inspector.
What does the customer gain, and what do they see?
They hear from you what is coming before their machine stops and can put the visit into a quiet period. In the portal they can see the condition of their own machines if you wish, never those of other customers. Messages to them go out only after service has approved them.
