Automotive and Manufacturing

Sovereign intelligence for Industry 4.0: Where visual precision meets agentic AI.

In an era where software-defined vehicles (SDV) and global supply chain disruptions set the pace, DEON transforms your workspace from a passive work surface into an active strategic partner. As a universal data aggregator, DEON connects isolated data silos – from CAD models and ERP live data to complex R&D documentation – on an infinite, high‑performance canvas.

With the new AI integration DEON Holistic Intelligence we provide professionals across the entire value chain with a tool that reduces cognitive load, stops the massive demographic knowledge loss, and radically accelerates decision-making through agentic support. We put the human back in the driver's seat by maximizing the range of interaction between human expertise and artificial intelligence, rather than hiding it in a black box.

Application areas of DEON across the industrial value chain

DEON covers the entire breadth of industrial core processes and offers specific solutions for various departments:

Research and Development (R&D): The platform serves as a virtual engineering space where teams collaborate on digital twins, embed live drone streams into plant layouts and annotate high‑resolution simulation data.
Product Development Process (PEP): Benefits from seamless integration of Microsoft 365, Jira and Azure DevOps, enabling requirements and design drafts to be visually linked and synchronized without redundancies.
Operations and Shopfloor Management:
AI-powered Knowledge Management:
Supply Chain and Procurement:
Quality Assurance: Supported by visual fault remediation, where engineers via viewport analysis link technical drawings with live inventory data from SAP to minimize downtime.

Use Cases

Knowledge preservation in the 'Knowledge Capture Race': Stopping the demographic brain drain

Challenge

The automotive and manufacturing industry faces an existential challenge: the massive outflow of decades-long experiential knowledge due to the retirement of the baby-boomer generation.

When experienced engineers and technicians, who have finely tuned production lines for decades, leave the company, the valuable implicit knowledge – that ‚diagnostic intuition‘, which is not found in any manual. Conventional methods such as text logs or confusing folder structures fail to capture complex visual and procedural relationships.

For successors it is often impossible to promptly dig through mountains of files and unordered email threads to understand the logic behind past decisions or technical setups. This skill gap becomes a bottleneck for overall operational excellence and endangers the productivity of entire sites.

Solution

DEON solves this problem with a visual and AI-powered knowledge transfer approach. Departing employees ‚empties‘ their knowledge unstructured on a DEON board – from email screenshots and handwritten sketches to process diagrams and embedded live documents.

The DEON AI kernel analyzes these visual relationships semantically and recognizes the layout like a human. Successors no longer have to work manually through documents, but simply ask the assistant DAIA simply:

‚What are my specific tasks on line X?‘
‚What was this test setup intended for?‘

The AI provides precise answers and uses visual grounding: Through interactive reference buttons, the workspace instantly zooms to the exact location on the board – for example into an embedded photo of a test bench or a specific Jira ticket.

This transparent approach prevents ‘hallucinations’, builds trust in the data foundation and enables an intuitive, verifiable knowledge transfer that reduces the onboarding time of new specialists by up to 60% reduced and persistently secures the corporate memory.

Multimodal technical analysis: Engineering acceleration in R&D

Challenge

Engineers in technical development today work in an extremely fragmented tool landscape. Critical information about a component or an experimental setup is distributed across various silos:

High-resolution CAD screenshots are stored in local folders
Excel bill of materials on SharePoint
Technical specifications in PDFs
Current market pricing or availabilities in web databases such as SAP Fiori or Google Sheets

Manual consolidation of this information for a solid technical analysis or a design review is time‑consuming and error‑prone. Often the overview is lost when technical details are discussed in linear chat threads or endless email chains, while the reference to the actual visual object – the drawing or the prototype – is constantly interrupted.

This leads to delayed innovation cycles and unnecessary errors in the early phase of product development.

Solution

DEON acts here as a universal data aggregator and visual analysis layer. Engineers can zoom into technical drawings or photos of experimental setups and ask the AI via the Viewport analysis (‚Eye‘-function) ask: ‚Explain this setup based on the visible electronics‘.

Thanks to multimodal capabilities, the AI understands the visual image including diagrams and sketches. At the same time, DAIA accesses data from live embedded web browsers. If the engineer asks, for example, about the cost of a specific component (e.g., ‘AS5600’), the AI extracts the price in real time from the embedded Google spreadsheet or the SAP interface and links it directly to the technical drawing.

Graph Chat and Visual Branching

Complex discussions are conducted over the ‚Graph Chat‘ as a visual tree on the board. Engineers can open separate branches (Branching) for various aspects such as ‘Toxicity’, ‘Cost’ or ‘Patents’, without diluting the main context of the discussion. This ‚Prompt Hygiene‘ ensures the highest answer quality.

The results are persistent DEON objects that can be collaboratively finalized and exported directly as editable PowerPoint for board presentations – a massive efficiency gain for the entire engineering workflow.

Digital Shopfloor Management: The 'Self-Building' Situation Center

Challenge

In modern production environments, overwhelming amounts of data are generated daily:

Hundreds of error tickets
Maintenance reports
Shift logs
Post-its from brainstorming sessions

Shift supervisors often spend hours manually sorting, prioritizing, and preparing this information for regular communication. The traditional analog shop floor board reaches its limits here, as it cannot process live data and does not involve remotely participating experts (e.g., from maintenance or engineering).

In dynamic situations, such as a sudden halt of the production line, valuable time is lost searching for logs, opening dashboards, and understanding the current architecture status. The gap between automated background processes and human decision-making leads to costly downtime and inefficient meetings.

Solution

With DEON, shop‑floor management becomes a self‑updating situation centre. Through the Model Context Protocol (MCP) the AI can actively control tools in DEON and manipulate content. An n8n workflow monitors the production systems and automatically pushes critical tickets to the board.

A production manager now simply issues the command: 'Group all error reports from the painting line and colour critical bugs red'. The AI semantically analyses the texts and automatically performs formatting as well as spatial arrangement.

War Room Mode

In the event of a machine failure, DEON acts as a 'War Room':

The AI highlights the affected component in the schematic architecture diagram
Pulls the last 50 error logs as text cards beside it
Opens a web browser in parallel with the live status dashboard

When the team enters the virtual room, the diagnosis is visually prepared. Through the Jira integration, structured actions can be synchronized directly back into the ticket system.

This saves valuable time in operational meetings, creates immediate visual clarity on critical topics and minimizes reaction time to disruptions through seamless Human-Agent Collaboration.

Claim Management & Compliance in Large Projects

Challenge

Large projects in plant and vehicle construction are characterized by massive change order claims and strict compliance requirements. Project managers often face the problem of having to prove who approved which change when, in order to fend off unjustified claims.

However, the relevant evidence is buried in thousands of logs on SharePoint, informal agreements in MS Teams chats, and countless emails. In internal audits or compliance investigations (e.g., when suspecting policy violations), departments must reconstruct internal communication flows extremely quickly.

The constant switching between different M365 silos makes it difficult to recognize patterns and relationship networks between people and documents. In a linear chat history, complex analyses often lose the overview, as important intermediate results disappear in the scroll history.

Solution

DEON solves this complexity problem through the synergy of Microsoft Copilot and visual structuring. The project manager uses DEON as a strategic dashboard and instructs the integrated Copilot: ‚Search all mentions of delay component X in the logs and emails of Q3‘.

By grounding on the Microsoft Graph, relevant information is extracted and visualized in DEON as native, editable objects.

Visual Evidence Chains

Instead of a ‘text desert’, visual evidence chains are created: The user drags the findings onto a timeline (Timeline) in the workspace and visually connects the email commitment with the later blueprint. Contradictions are examined in the ‘Graph Chat’.

Through ‘Visual Branching’ specific aspects of a suspicion case (e.g., an email attachment) can be deepened in a separate branch, without losing the strategic thread of the overall analysis. Every statement from the AI remains 100% traceable to the original source via visual links.

The result is a persistent, visual knowledge map that reveals complex relationship networks and can be directly transferred into legally robust documentation. This ensures full data sovereignty within the protected corporate space (on-premises or sovereign cloud).