Why AFFERENT exists
Modern life rests on machines that almost no one sees. Pumps move water. Chillers keep servers within their limits. Transformers step voltage down for a neighborhood. Motors turn the lines that make nearly everything. This equipment runs for years with very little attention, and it is expected to keep running.
When it stops, the cost arrives at once: a production line idle, a data hall overheating, a district without power. And in a great many cases the equipment gave warning. It ran a little hotter, vibrated a little differently or made a sound it had not made before. The warning went unnoticed because nothing was there to notice it.
AFFERENT exists to close that gap. Computing has learned to interpret language, images and software. The physical systems that support daily life remain, for the most part, unable to explain their own condition. We think that is the next important problem for computing to take on.
The nervous-system analogy
The company's name describes its architecture. In anatomy, afferent pathways carry signals from sensory receptors toward the central nervous system. They are how a body knows what is happening at its own edges.
A building, a factory or a power network has no such pathways. It has controls, which tell machines what to do, and very little sensation, which would tell operators how the machines are. AFFERENT builds the missing half. NODE devices act as receptors. Networks carry their signals. AFFERENT Cloud is where signals are combined into awareness.
What we build
NODE
An intelligent device that mounts on machinery and infrastructure, designed to observe an asset through several kinds of sensing and to analyze what it observes at the edge. NODE is planned as a family: Mini, Industrial and Critical.
AFFERENT Cloud
Software designed to bring together what devices report, maintain the history of each asset and present machine health in a form people can act on.
Machine Health Intelligence
The models, and the accumulated understanding behind them, that are intended to recognize how equipment behaves before it fails. Read the thesis.
Long-term vision
Predictive maintenance is where this work begins, because it is where the value is most immediate and most measurable. It is not where the work ends.
If physical systems can be observed continuously and understood reliably, a good deal else follows. Equipment can be operated closer to its real limits instead of conservative assumptions. Maintenance can be planned around actual condition instead of the calendar. Designers can learn how their machines age in service. Infrastructure that spans a region can be managed as a single system whose state is known.
We describe that future as a physical world that has become observable. It will take many years and a great deal of careful engineering. We intend to build toward it one verified capability at a time.
Principles
- Observe first
- Understanding begins with sensing. We do not theorize about a machine we could measure.
- Intelligence at the edge
- Critical decisions should not always depend on distant computation. The device must be useful on its own.
- Design for reality
- Physical environments are messy. Technology must survive heat, dust, vibration and the absence of a network.
- Earn trust
- Infrastructure intelligence must be secure, explainable and dependable. We state what is true and label what is planned.
- Think in decades
- We are building architecture meant to become infrastructure. Decisions are judged by how they will hold up over a long service life.
How we describe ourselves
AFFERENT is a new company with a large ambition. On this website we try to keep those two facts in proportion. Product pages state their current status. Illustrative graphics are labeled. We do not publish customer names, deployment figures, performance statistics or certifications that we cannot support with evidence.
Credibility is part of what we are building. In this industry it is also the thing customers buy first.