Dispatch is usually treated as a logistics problem.
In most organizations, it is framed as a scheduling task: who is available, how quickly they can get on-site, and how to keep queues moving. That framing is convenient, but it obscures where reliability outcomes are actually determined.
Dispatch is not clerical. It is a high-consequence decision made under uncertainty. By the time a technician is rolling, much of the outcome—success, rework, delay, or escalation—has already been shaped upstream.
Dispatch as a Decision Authority Under Constraint
Every dispatch implicitly answers questions that go far beyond calendars and routing. Is the outage real or transient? Can it be resolved remotely, and if so, is that intervention safe? If field service is required, who should go, with what scope, under what safety posture, and within what site access constraints?
These questions live squarely in Tier 2.
That does not mean Tier 2 operates with sovereign control. Tier 2 does not own dispatch schedules, site host availability, utility coordination, or contractual permissions. In many cases, it can’t even issue a remote station reset. But Tier 2 does bear responsibility for identifying and articulating the constraints that shape dispatch outcomes—particularly site access windows, site access restrictions, required credentialing or PPE, and dependencies that can quietly turn a well-intentioned truck roll into a failed visit.
Tier 2 operates as a decision authority under constraint, not a sovereign control layer. It is still expected to decide, and it is still expected to be right.
The Failure Mode of “Just Send a Truck”
When diagnostic confidence is low, many systems fall into a familiar pattern. An alert fires. The fault description is vague. The cost of delay feels higher than the cost of action. So the system escalates—often by sending a senior technician “just in case,” with a loosely defined scope and the expectation that the field will figure it out on arrival. In the early days at EV Connect, I did this—a lot—and the cost in time, run-around, and stamina was significant.
When this approach fails, the failure is often attributed to execution: the wrong technician, the wrong part, unclear access, or the need for a second visit. In reality, these are not technician failures. They are system failures that substitute escalation for diagnosis and hope for preparation.
Over time, this pattern reshapes the workforce. Senior electricians become the default shock absorbers for uncertainty. Junior technicians are underutilized. Repeat visits normalize, and the best technicians get overloaded. Truck rolls begin to feel scarce—not because labor is insufficient, but because ambiguity is being pushed downstream.
When Tier 2 lacks confidence, Tier 3 absorbs uncertainty.
First-Time Fix Is Decided Before Arrival
First-time fix is often treated as a field metric. In practice, it is a decision-quality outcome.
Successful first-time fixes depend far more on upstream preparation than on downstream execution. Diagnostic clarity, scope definition, safety classification, parts readiness, and site access coordination all shape the likelihood of success long before a technician arrives on-site.
Field execution rarely fails on skill. It fails on preparation.
By the time a technician pulls into the site, the system has already made a series of decisions that determine whether success is likely or whether rework is inevitable.
What We Mean by the “Dispatch Stack”
This is where the idea of a dispatch stack becomes useful.
In many organizations, dispatch is implicit. Alerts turn into tickets. Tickets turn into truck rolls. Escalation substitutes for certainty, and confidence is assumed rather than assessed.
The dispatch stack reframes dispatch as a layered decision process rather than a moment in time. At a minimum, that process includes signal intake, correlation and context, confidence assessment, scope definition, and explicit dispatch or no-dispatch authorization.
Dispatch is not a moment. It is a stack of decisions, and weaknesses anywhere in that stack propagate forward.
That stack, in practice, is only as strong as the context available to Tier 2.
Correlation and Context Under Partial Visibility
Modern DC fast chargers generate far more diagnostic data than most networks ever see. What reaches Tier 2 is shaped by what EVSE manufacturers choose to expose through OCPP, and that exposure varies widely. Some manufacturers surface rich component-level telemetry. Others expose abstracted fault states, oversimplified descriptions, or numeric error codes that require interpretation and experience to decode.
This is not malice. It reflects uneven maturity, architectural history, and differing design priorities.
The practical consequence is that Tier 2 is often asked to correlate signals using partial telemetry, abstracted fault descriptions, historical patterns, site context, and experience. Direct access to deep charger telemetry is rare. Decisions are made through inference—sometimes supported by tribal knowledge, and occasionally by reaching out to EVSE manufacturers when personal relationships exist.
Tier 2 makes consequential decisions with partial visibility not because data does not exist, but because it is inaccessible at the moment decisions must be made.
Technician Insight Completes the Stack
This is where skilled technicians add irreplaceable value.
Experienced field technicians recognize patterns that do not show up cleanly in data streams: environmental contributors, installation-specific quirks, and recurring failure modes across sites and vendors. When that insight flows back upstream—through structured notes, feedback loops, and shared pattern recognition—it improves Tier 2 confidence over time.
The dispatch stack improves not by eliminating human judgment, but by learning from it.
Where AI Fits
AI helps quietly, and in the right place.
Its value lies not in automating dispatch, but in compressing uncertainty within the stack. It can correlate imperfect signals across time and sites, surface likely root causes versus symptoms, and score confidence in remote remediation versus field escalation.
Used this way, AI doesn’t automate dispatch—it makes dispatch more deliberate.
