MachineCopilot · Physical AI by TECHTORY

The autonomous machine assistant. Local AI for machinery and production.

MachineCopilot brings AI right to your equipment: in automation systems, at production machines or as a maintenance assistant. Your team asks by voice or touch, and the machine answers from the manual, fault history and live data, always with a source. The AI runs locally in your plant, without a cloud connection.

Or reach us directly: +49 7805 9589-0·info@techtory.de

4Business units
∞Possible combinations
140+Employees
30+Countries

The problem

The most expensive part of downtime is the search.

Whoever fixes a fault spends most of the time not with a spanner but searching: leafing through the manual, hunting for the fault list on a network drive, calling the colleague who had the same problem before. The knowledge that would shorten the search exists. It is just spread across three places that never meet.

Knowledge is scattered

Operating manual, wiring diagrams, spare parts lists: as PDFs on a network drive or in a binder at the control cabinet. Alarm history and operating data sit in the controller.

Knowledge retires

The setter who knows why a program runs at 80 % override will not be there forever. What was never written down leaves with them.

The expert is not there

Saturday, 2:30 a.m. The line is down, the expert is off, and the manufacturer’s support is not available until Monday.

Capabilities

What can AI do in your production?

TECHTORY AI develops data-driven and AI-based functions for machines and production processes, as an extension of industrial production systems. At its core is the autonomous machine assistant: for automation systems, at production machines or as a maintenance assistant.

How MachineCopilot works →

What MachineCopilot can do

See. Understand. Analyze. Act.

We bring AI to where your value is created: to the machine. Not as a promise for the future, but as a working system with four capabilities that build on each other.

Running on our own machines

Understand: ask your machine

Everything known about the machine becomes accessible by voice: manuals, fault history, operating information and live data from the controller. Every answer cites its source, down to the page in the manual.

„What does this alarm mean, and what do I do first?“

Running on our own machines

Analyze: causes instead of error codes

The system combines machine status, process data and hands-on experience and names the probable cause of a fault, with its reasoning. The maintenance technician makes the decision.

„Axis 2 reports a following error, but only in rapid traverse. Mechanics or drive?“

Shown live: aaa Zurich 2026

See: recognise objects without teaching

The machine captures its surroundings by camera, identifies objects and their position and passes the data straight to the robot or controller.

Picking without teaching in the Vision AI cell.

Next stage

Act: operation on voice command

Selected functions are operated by voice command, based on a tested and approved command catalogue. The AI does not improvise; emergency stop and safety circuits remain in the safety system.

„Open the drawers.“

All capabilities in detail →

Your benefit

Results that reach the shop floor

In the end, what counts is not the technology but the result at the machine.

Maintenance and service

  • Get to the cause faster: manual, history and current alarm in one answer.
  • Spare part number and wiring diagram reference in seconds instead of leafing through binders.
  • Maintenance log by voice, hands stay on the tool.

Production

  • Shorter downtime on night and weekend shifts.
  • New operators get up to speed faster: they ask the machine instead of the setter.
  • Guided changeovers from the setup sheet, no matter who is on shift.

Management and machine builders

  • Hands-on experience stays in the company, even through retirement and staff turnover.
  • Data and process know-how never leave the plant.
  • A predictable start: a pilot with fixed scope, fixed price and measurable goals.

Where is the benefit greatest on your machine? Let us find out →

Proof

Shown, not just described

There are enough AI promises. That is why we show what works and say openly what is still the next stage.

TECHTORY booth at all about automation Zurich with visitors at the Vision AI cell

Live at all about automation Zurich 2026

At booth G38 our Vision AI cell ran non-stop for two days. In the Talk Lounge, the presentation „The autonomous machine assistant“ showed the concrete fields of application.

The Vision AI cell with control station

Use case: Vision AI cell

Recognise, measure and pick arbitrary objects without teaching and without fixed fixtures. The entire AI runs locally in the cell. The interactive view explains the cell station by station.

By machine builders, for machine builders

We are not a software house that has discovered industry. TECHTORY has been building machines and automation solutions for more than 30 years, with 140+ employees in Appenweier, Germany. The machine assistant runs on our own equipment before it comes to you.

„We don’t know machines from slides. We build them ourselves.“

Edge AI

Why does the AI run locally and not in the cloud?

Because drawings, process parameters and fault histories are trade secrets. MachineCopilot runs on a compact edge unit right at or next to the machine. No external data connection is required.

✓

Data sovereignty

Your production and machine data never leave the plant. Auditable for IT security and data protection.

✓

No cloud approval

The edge unit sits on the machine network, not on the internet. There is no cloud connection your IT would have to approve.

✓

Available when the fault occurs

Response time does not depend on the internet. Even on Saturday at 2:30 a.m.

✓

The PLC stays the PLC

The assistant reads controller data, it does not write to the controller. Emergency stop, safety circuits and real-time control remain in the certified safety system.

What this means for your IT →

MachineCopilot pilot

Your own pilot in 4–8 weeks

One machine family, a clearly defined use case, fixed scope and fixed price. At the end you know, based on agreed metrics, whether MachineCopilot pays off for you.

  1. Pilot consultation · 30 minutesWe work out which machine offers the greatest leverage. No obligation, with a live demo at your site on request.
  2. Weeks 1–2 · DefineOn-site workshop: choose the machine family and use case, collect documents, clarify interface and success metrics.
  3. Weeks 3–6 · Build and testEdge unit at the machine, documents loaded, controller data connected read-only, first questions from your team checked.
  4. Weeks 7–8 · Measure and decideUse in shift operation, evaluation of the metrics, recommendation for next steps.

Let’s talk about your machine

In the pilot consultation we work out where the machine assistant has the greatest leverage in your production. Concrete, honest and on equal terms.

Or reach us directly: +49 7805 9589-0·info@techtory.de

Fixed scope · fixed price · we usually reply within one working day

FAQ

Frequently asked questions about MachineCopilot

Does our data leave the plant?

No. MachineCopilot runs locally on an edge unit in your plant. No cloud connection is required.

Does the AI interfere with our controller?

No. The assistant reads controller data via standard interfaces; it does not write to the controller. The safety system remains untouched.

Does it work on older machines?

Yes. MachineCopilot is vendor-neutral and can be retrofitted. Which data your controller provides is clarified in the pilot.

All questions and answers →