Manufacturing
Most of the useful signal is already in the equipment
In a plant, unplanned downtime is the expensive event. The readings that would predict it are often already being produced by machines that do not readily share them.
Context
Industry context
What shapes technology decisions in this sector.
Manufacturing is a monitoring and integration problem before it is an analytics one. Machines produce signal; the difficulty is getting it out of equipment that was not designed to share it, and into a form operations can read.
Maintenance is usually scheduled or reactive. Both leave a gap: a fault developing between intervals is found by failure, and the cost of that failure is rarely limited to the part that broke.
Energy and output are frequently measured only in total. Per-line and per-machine figures change what can be attributed and therefore what can be improved.
Constraints
Key operational challenges
The conditions any solution here has to work within.
- Machines that fail between scheduled maintenance intervals.
- Production data held in equipment that does not readily share it.
- Energy use per line or machine that is not separately measured.
- Older equipment that has to be instrumented without being replaced.
Capabilities
Relevant ZouhThings capabilities
What we can bring to this sector, stated as capability rather than track record.
Condition monitoring of assets
Can support continuous measurement of temperature, vibration, current and running hours as indicators of equipment condition.
Operations visibility
Designed for bringing production, energy and equipment status into one operational view rather than several local panels.
Integration with installed equipment
Relevant where older machines have to be instrumented without being replaced, using whatever they expose.
Data and analytics
May enable identifying patterns that precede a failure, and separating a developing fault from normal variation.
Cybersecurity for connected assets
Designed for keeping a monitoring layer separated from production control, so visibility does not become a path to actuation.
Solutions
Relevant solutions
Solution areas that apply here. Each links to how it is delivered.
Platforms
Relevant platforms
Platforms that would apply to this sector.
- thethings.io
Applications
Example use cases
Possible applications, to show the shape of the work.
Predictive maintenance on a critical asset
Possible use cases include instrumenting one critical pump, motor or line and alerting when readings leave a normal band.
Per-line energy attribution
Can support measuring consumption per line or machine, so energy cost is attributable rather than a site total.
Production and downtime visibility
May enable recording downtime with its cause, so the largest recurring loss is identifiable rather than anecdotal.
Spare-part planning from condition data
Designed for informing inventory from observed condition rather than from a fixed schedule.
These are possible use cases, described to illustrate what the technology can support in this sector. They are not descriptions of completed work, and none should be read as a reference. Where a confirmed project exists, it will be named and linked under Related projects.
Security
Security considerations
What has to be decided early rather than added later.
Monitoring separated from control
Relevant because a monitoring layer must not be able to actuate a machine; segmentation is a design requirement rather than hardening.
Identity for connected assets
Each connected device identifiable and revocable, so instrumenting a machine does not create an unmanaged endpoint.
Protecting production data
Can support keeping production and process data confidential, which is often commercially sensitive in its own right.
Update paths that respect uptime
Designed so devices can be updated without stopping a line, decided before deployment rather than after.
This describes how we approach security in this sector. It is not a claim to hold any specific certification, accreditation or compliance attestation — where one exists, it will be named.
Integration
Integration considerations
Where the difficulty usually sits in this sector.
Older equipment first
Designed for instrumenting machines that predate the project, without requiring their replacement.
Protocol and format differences
Can support normalising differing protocols, units and intervals into something one view can present.
Behaviour when the network drops
Local buffering with observed and received times separated, so a backlog is not read as a burst of faults.
Alerting the maintenance team can trust
Relevant because an alert volume that gets muted removes the only benefit the instrumentation was bought for.
Evidence
Related projects
Confirmed work in this sector, where any is published.
No project is published for this sector yet. When one is confirmed and approved for publication, it will appear here — so an empty section means exactly that, rather than work we cannot discuss.
Next step
Discuss a requirement in this sector
The useful conversation is usually about a specific asset, site or corridor rather than the sector in general.