中文简体A swapping cabinet showing an empty slot only means that slot door is empty; whether a rider can actually swap for a charged battery depends on whether there is a fully charged, healthy battery in the slot that the cabinet can recognize normally. These are two different metrics: the number of empty slots looks at slot occupancy, while the number of available slots looks at service capacity. How many times a 10-slot battery swap cabinet can serve during the two-hour evening peak is determined jointly by the number of batteries, the full-charge time of a single battery, and the length of the peak window—not by "how many empty slots are left." Looking at these three numbers together is what creates the deadlock where "the backend says there are slots, but riders say there are no charged batteries."
Empty slots, fully charged slots, and available slots are three different numbers
CompartmentIndicator lightTwo of them can be seen directly: a light that is off means an empty slot, and a green light steady on means fully charged. A red light steady on is an intermediate state—that battery is charging, and it cannot be taken even when someone arrives. As for "available slots", the light cannot show it; it must be judged by three conditions together.
The slot indicator lights on Dao Chong battery swap cabinets distinguish five states: steady green means the battery is fully charged, steady red means charging, no light means an empty slot, flashing yellow means a battery is being inserted (some models), and both red and green on means the charging slot is faulty (some models). Both green and red indicate that there is a battery in the slot, and in a backend that only looks at occupancy, both count as "occupying a position," but they mean completely different things to riders—for the red-light slot, even if a rider arrives, they cannot get a battery.
A slot that can truly serve customers must meet three conditions at the same time:Battery fully charged, battery healthy, cabinet online. If any one of these does not hold, the compartment exists but the power cannot be drawn.
Status seen on site | The dashboard is easily misread as | Can it actually serve? |
|---|---|---|
Light off (empty slot) | Has an empty position | No, there is no battery |
Red light steady on | In stock | No, the battery is charging |
Flashing yellow light | In stock | No, a battery is being inserted and its status is undetermined |
Red and green both on | Device online | No, this compartment has reported a fault |
Green light steady on but battery aging | Has a fully charged battery | Can be swapped, but riders will feel it "doesn't last long enough" |
Whole cabinet offline | Light still on | No, status is no longer being uploaded |
A dashboard that only looks at the "number of empty slots" is seeing a distorted picture.Many empty slots may mean not enough batteries; few empty slots may also mean batteries are stuck charging.
How many swaps can be done during peak periods can be calculated
Battery swapping is a business of "concentrated demand + continuous supply": demand is squeezed into a one- or two-hour window, while supply is determined by charging time—the rhythms of the two are naturally out of sync.
The algorithm is not complicated. After a battery is swapped out, it must go through a charging period before it becomes an available asset again. Within a peak window of length H, the upper limit on how many times a battery can be swapped out is
⌊H ÷ T⌋ + 1(T is the time required to fully charge one battery, ⌊ ⌋ indicates rounding down, and the +1 at the end is "the fully charged one in hand at the start").This formula gives a count, not an average rate. 2 ÷ 1.5 comes out to 1.33, but that is a rate, not "1.33 times"—a single battery can only be swapped 1 time, 2 times, 3 times. If this step is not rounded, supply capacity will be overestimated, and the gap will instead be narrowed.
Take a Dochion 10-compartment cabinet with 10 batteries as an example. The maximum power of the standard 10-compartment whole cabinet is 8kW, which works out to about 800W per compartment; at 90% conversion efficiency, about 720W per compartment actually goes into the batteries. Assume further that the batteries are the common 48V／20Ah specification used on e-bikes (nominal capacity 48×20＝0.96kWh); going from 20% to 100% requires about 0.77kWh to be charged in, which at a rough 720W takes about 1.07 hours; in practice, current tapers off after charging enters the constant-voltage stage, so 1.5 hours is closer to what is seen on site.
The whole-cabinet 8kW in the previous section is the value listed in the parameter table; the per-compartment 800W is derived as 8kW ÷ 10 compartments (the parameter table does not list per-compartment power separately); battery capacity and charging time are likewiseExample values. Actual values are subject to the cabinet nameplate, battery model, and BMS charging strategy.
Calculated based on a 2-hour peak and 1.5 hours to fully charge:
Project | Value | Description |
|---|---|---|
Number of times a single battery can be swapped during peak hours | 2 times | ⌊2 ÷ 1.5⌋ + 1 |
Peak supply capacity of 10 batteries | 20 times | upper limits, not accounting for faults, dispatch losses, and riders' time using batteries outside |
Actual peak demand at this site | 40 times (assumed) | Concentrated battery swapping by nearby riders |
Gap | 20 times | Manifests as "there are empty slots but no swap can be made" |
Once a gap appears, adding batteries is only one solution, and it is not necessarily the more cost-effective one. The same gap can be filled either by adding batteries or by shortening the turnover time of each battery—the latter adds no assets, but it depends on whether the cabinet's power and charging strategy allow it.
When a gap appears, check four places first
Before deciding to add assets, first recover the "capacity that should already exist." From what can be seen on site, four categories of causes eat into availability, and all of them come before "buying new batteries":
- How many batteries are not in the pool —— Batteries under repair, at another site, or on the books but not physically present all count as lost throughput, and such losses do not show up in the empty-compartment count.
- Is it actually charging? —— A charger fault or a compartment lock fault can make a compartment appear to hold a battery while it is not actually charging; the light behaves much like "charging".
- Whether scheduling was done during peak periods —— Fully charged batteries pile up at off-peak sites while peak sites can only wait; this is a dispatch problem, not a quantity problem.
- Has the whole cabinet ever been offline? —— A single offline period of several tens of minutes means zero throughput at that site during that time, yet the indicator light in the back end stays on.
Among these four categories, troubleshooting the first three depends on "whether you can see where the batteries are"—batteries are mobile assets, and their location and status must be traceable so that scheduling has a basis; the cabinet-side water immersion detection automatically stops charging after detecting water ingress, preventing "charging while faulty" from being treated as normal turnover.Only what can be seen can be scheduled.
Which metrics to watch
Operations should watch not the number of cabinet doors, but the number of fully charged assets and their turnover speed during peak periods. The five indicators below: the first three show "whether there is power to swap right now, and whether it is enough"; the last two show "why it may not be enough":
Indicators | How to calculate | What it indicates |
|---|---|---|
Availability rate | Number of slots with fully charged and healthy batteries ÷ total number of slots | The service level of a site at this moment |
Peak gap | Peak demand count − Peak supply count | Whether to expand, and by how much |
Average daily turnover per slot | Daily swap count ÷ number of slots | Whether assets are utilized fully enough |
Proportion of faulty slots | Number of faulty slots ÷ total number of slots | How many slots are occupied without producing output |
Number of offline batteries | On the books, location unknown, or not returned to the cabinet for a long time | How many assets are leaking |
Looking at it the other way makes it clearer:The service capacity of a cabinet is "number of fully charged batteries × peak turnover count"; none of these terms is called "number of compartments". The number of compartments determines the upper limit, not the amount that can be delivered on a given day.
This also explains why the same cabinet placed at different locations can produce very different results—it is not the difference in cabinets, but the different combination of three variables: the number of batteries, the charging duration, and the length of the peak period. To judge whether a location needs additional assets, the order should be: first calculate the peak gap, then check whether existing assets are being wasted, and only then discuss procurement.