Miles of Infrastructure That Fails Silently

An eruv is a continuous boundary, usually a thin wire carried over a series of vertical posts, that lets an observant community carry outdoors on Shabbat. A mid-size American city eruv runs 8 to 25 miles and touches hundreds of utility poles, and it is valid only if every span is intact. One snapped line anywhere along the route invalidates the whole enclosure, and nothing about the failure announces itself.

The traditional answer is a weekly inspection: someone walks or drives the entire route on Friday morning, checks every attachment with binoculars, and reports before candle lighting. That is four to eight hours of skilled volunteer labor every week, in every weather, against a hard deadline. It is the clearest case in community infrastructure for remote monitoring.

What There Actually Is to Sense

The failures worth catching are narrow, which is good news for the sensing design. A span breaks or sags below the post top. A post is knocked out of plumb by a vehicle or storm. A utility crew removes an attachment during line work and does not put it back. Vegetation grows through a span. Ice loading pulls a run down.

Three sensing approaches cover almost all of it, and most deployments mix them:

  • Tension or load sensing at the anchor. A small load cell or spring-and-switch assembly at the terminating post reads the pull of the run. A break drops the reading to zero within a second. This is cheap, unambiguous, and covers every span between two instrumented anchors, which is why it is usually the backbone of a system.
  • Electrical continuity. If the boundary material is conductive, or if a fine conductive strand is run alongside a nylon line, a loop current gives you a binary answer for a whole segment with one node at each end. Watch for corrosion at splices and for the loop resistance drifting with temperature over a long run.
  • Vision at the hard points. A low-power camera node aimed at a specific attachment, comparing a fresh frame against a stored reference, catches the failures the other two miss, such as a line that is intact but has slipped off the post. Do the comparison on the node and send a small confidence value plus an occasional thumbnail, not a video stream.

Post lean comes almost free: the three-axis accelerometer already on the node board works as an inclinometer, so alarm on a few degrees of change from the installed baseline.

Power and Communications in the Field

Nodes hang on utility poles with no service access and no mains. That sets the whole electrical design: a small solar panel, a lithium iron phosphate cell chosen for cold-weather charge behavior rather than energy density, and a duty cycle measured in reports per hour instead of readings per second. A node that samples tension continuously but transmits four times a day will run for years. The tradeoffs, including when a panel beats a primary cell outright, are worked through in energy harvesting for IoT devices.

For the radio link, an eruv is the textbook case for narrowband cellular: the nodes are spread over a wide urban area, each sends a few dozen bytes, and nobody wants to own gateways. Compare the carrier options against your coverage maps using LTE-M vs NB-IoT. If the route is dense and one organization controls a tall building or two, a private long-range network can cut per-node subscription cost to nothing, which is the calculation in LoRa and LoRaWAN in a product.

Building Something That Survives Outside

Pole-mounted hardware sees everything: driving rain, road salt spray, ultraviolet exposure all day, thermal cycling from below zero to well above 120 F inside a dark enclosure in summer sun, and the occasional bird. Specify the enclosure to a sealing rating you can actually verify and understand what the test does and does not prove, per IP ratings explained. Then pick the housing polymer for years of direct sun rather than for molding convenience, because a housing that chalks and cracks in three summers turns a maintenance-free product into a truck roll, a failure pattern detailed in UV-resistant plastics for outdoor products.

Two practical constraints catch first-time developers. Attaching anything to a utility pole requires a joint-use agreement with the pole owner and compliance with electrical clearance rules; the hardware must be installable by an approved crew, not by a volunteer with a ladder. And the node has to be serviceable in ten minutes from a bucket truck, which means captive fasteners, a battery that swaps without tools, and a status indicator visible from the ground.

The Software Is the Product

Hundreds of nodes on a schedule is a fleet problem before it is a sensing problem. You need per-node health, battery and solar trend lines, firmware rollout in stages, and alarm deduplication so a storm does not generate two hundred pages. That discipline is the subject of IoT fleet management, and skipping it is the most common reason a pilot of twenty nodes never scales to a full route.

Alarm timing should be built around the community's real deadline. A break detected Wednesday afternoon leaves two full days to schedule a repair; the same break detected Friday at noon is a crisis. Escalation paths, a maintenance window before the weekly deadline, and an offline node treated as a fault rather than as silence are all part of the spec.

The System Reports, a Human Rules

This is the design constraint people outside the market miss. The monitoring system never declares the eruv valid. It reports physical facts: this span holds tension, this node has not checked in, this attachment looks different from its reference image. A person inspects what the system flags, and a rabbi decides. Build the interface that way, with plain sensor language and an explicit human sign-off step before any public status is published, and keep an immutable log of who signed off and when. The same separation between measurement and ruling governs the camera and sensor systems described in remote kosher supervision technology.

Once the network exists, the community owns a wide-area outdoor sensing platform, and the marginal cost of adding facility monitoring for the buildings inside the boundary is small.

Building an Eruv Monitoring System

Projects House develops outdoor sensor products end to end: sensing approach, node electronics, solar and battery sizing, enclosure, and the fleet backend behind them. If you are a community organization or a supplier looking at this market, describe your route length and node count in the contact form and we will scope a pilot.