Skip to content
AFCONX

Projects / Internal tool

TGW Insights

One platform for the maintenance team of an automated fulfilment centre.

All data shown is anonymised. The final software may differ from what is presented here.

TGW Insights is a product of TGW Logistics Group GmbH. © 2026 TGW Logistics Group GmbH. All rights reserved.

Context
Internal tool at TGW Logistics
Year
2026
Role
Product owner, design, development
Status
Shipped
The Insights home dashboard with demo data: site health, today's handovers, watch list, maintenance and shifts

Why

What stopped this area last night, and has it happened before? Answering that took four systems and someone who knew where to look.

Before

Alarms in one system, planned work in another, parts in exports, handovers in notes, and KPIs assembled by hand.

The idea

One place that reads from the systems that already exist, so nobody types anything twice.

Today

Between 60 and 70 people use it every day, from technicians on shift to site management.

What the team gets

From the alarm on the wall to the part on the shelf.

Alarms you can read

What is standing right now, what happens most, and how one fault cascades into others.

The alarm timeline with demo data, one lane per controller, with a cascade marked

A board for the wall

You can read it from across the workshop, so nobody has to walk over to see how the site is doing.

Downtime and KPIs

The period reports used to be assembled by hand. Now they come out of rules anyone can read.

Planning and tasks

A calendar with my tasks and the team's, fed straight from the maintenance management system.

Parts and deliveries

It covers stock, order suggestions, receiving and labels, down to the shelf a part sits on.

Shift handover

What the next shift needs to know arrives in one stream instead of loose notes.

Alerts in Teams

Build your own alert recipes, and they arrive where the team already talks.

Built with

Self-hosted, inside the plant.

  • Next.js
  • React
  • TypeScript
  • PostgreSQL
  • Prisma
  • Docker
  • OpenTelemetry
  • Grafana
  • Modular architecture
  • Live data feeds
  • Scheduled jobs
  • Passkey sign-in
  • Role-based access
  • Teams alerts
  • Encrypted backups
  • Tracing, metrics and logs

11

modules in one app

145

pages

60–70

people using it

Design

What happened, where, and what to do about it.

Written for the floor

The systems underneath hold the kind of detail an analyst wants. Every screen turns it into what happened, where it happened, and what to do next.

Everything points at the equipment

An alarm, a task, a part and a handover note all hang off the same machine, so its full picture is never more than a click or two away.

One shell, eleven modules

Every module sits in the same frame, so a technician who learns one page already knows the other ten.

Real Insights screens, shown with anonymised data.

The Processes page with demo data: downtime per process, with health, trend and a sparkline
Processes: downtime per area of the site, in the floor's own terms.
The alarm timeline with demo data and a cascade marked
Timeline: one lane per controller, with a cascade marked in red.
The maintenance calendar's week view with demo data, tasks per shift and team
The week's maintenance, per shift and team, done work struck through.
The Shuttle Fleet page with demo data: fleet status, what needs someone, and a live feed
Shuttle Fleet: what needs someone now, and what just happened.
The Insights sign-in screen with passkey and email options
Sign in with a passkey, or with email.