Inventory

Why most lab inventories become fiction

Almost every electronic lab notebook ships an inventory, and almost every academic lab abandons the one it has. The reason is always the same. The system stores a precise running amount (for example “47.3 mL remaining”) and expects you to decrement it every time you pipette. Nobody does that for more than a week. Within a month the numbers are wrong, and within a quarter the whole inventory is fiction. A wrong inventory is worse than no inventory, because people trust it, skip a reorder, and find an empty bottle on experiment day.

So Inventory is built around a different question. The question is not what we can track. It is what a busy lab will realistically keep current with near-zero effort, and how to make those few things produce real value. The answer is to count containers instead of volumes, flip status with one tap instead of doing math, and let the passage of time compute the rest.

The big idea, count containers not volumes

The primary number on a stock is a count of physical containers. Three vials, two bottles, one plate, a box of tips. You change that count only when a whole container is finished (you throw out the empty tube, three goes to two) or a new one arrives. That is an event that happens roughly monthly per item, it is unambiguous (the empty tube is in your hand), and it takes one tap. It never asks you to remember how much you used, do arithmetic, or update anything mid-experiment.

There is an optional per-container amount label (for example “1 mL” or “100 ug”) for people who want to record how big each container is, but it is never required and it is never decremented for you. It is a label, not a ledger. A lab that wants more detail can add it, and a lab that wants none never sees a number it is obligated to keep accurate.

Alongside the count, every stock carries a coarse status (in stock, low, empty). You flip it with a one-tap control the moment you notice at the bench, because your eyeball is the sensor and no threshold math is needed for that to be useful. If you also set a low-at-count threshold (for example, flag when vials drop below two), the status flips itself and even the tap becomes optional. The point is that Inventory is useful whether you maintain a count, tap a status, both, or neither, and it degrades gracefully instead of lying.

The three zero-upkeep signals

The payoff of an inventory is not the stock list you scroll. It is three signals, and all three are derived from data you enter once at receive, or not at all, and then maintained by the calendar rather than by you. These are the center of the feature, not an afterthought.

• Expiring soon and expired. Driven by the expiration date you typed once. The clock does the rest forever. This is what keeps you from running a failed experiment on dead reagent, or tossing money on an antibody that lapsed months ago.
• Stale and untouched. Driven by the received date (stamped automatically) plus the last time the record was touched (also automatic). Pure time math, zero ongoing input. This surfaces the six forgotten tubes in the back of the freezer, and the inverse, a critical reagent nobody has confirmed is still there.
• Low or empty. Driven by the container count and an optional low-at-count threshold, or by a single manual low tap. This is what keeps you from re-buying what you already have, and from discovering you are out on experiment day.

None of these needs an accurate running total of anything. Expiry and staleness need zero ongoing input at all. That is the point. The value lives in signals that the passage of time computes for free.

The Supplies hub, where ordering becomes inventory

Inventory does not live alone. It shares a home with Purchases under a single Supplies hub, presented as two tabs. The reason they sit together is that one feeds the other. Purchases answers “what did we buy.” Inventory answers “what do we have, where is it, and is it still good.” The receive event is the seam between them.

When you mark a purchase order received, which is something the buyer already does, Inventory offers to add it for you. It pre-fills the name, vendor, catalog number, and link from the line item you bought, defaults the container count to the quantity ordered, stamps the received date, and asks only for expiry and (optionally) a location. The buyer is already standing there with the box, so capturing it costs one extra field. Over a few months of normal ordering, the inventory builds itself out of the reagents the lab actually churns through, with no separate data-entry chore.

The flow runs the other way too. When finishing a container drops an item below its low threshold, that item lands in the Purchases reorder queue, so “we are nearly out” turns into “this is on the list to buy” without anyone retyping the reagent. The loop closes (you buy it, you receive it, it becomes stock, you finish it, it queues to buy again) and each hand-off carries the details forward for you.

Barcode scanning from the desktop webcam

Finishing a container is a roughly monthly one-tap event, and a barcode scan is simply that tap made faster. ResearchOS scans through your desktop webcam (the app is Chrome and Edge only, so the browser's built-in barcode reader is available), no separate hardware required.

There are two things you can scan. A container code is a lab-applied label on one specific tube or box. Scanning it counts that container down by one, re-derives the status (to empty at zero, to low when it crosses your threshold), and drops the item into the reorder queue when it crosses low, all guarded by a quick-undo toast in case you mis-scanned. A product barcode is the manufacturer code shared by every container of a product. Scanning a known one finds the item, and scanning an unknown one starts registering it.

Lab reagents from suppliers like NEB, Sigma-Aldrich, and Thermo Fisher mostly do not carry the retail barcodes that consumer databases know, so the high-value path is the lab-applied container code you control, not a manufacturer lookup. An optional product lookup can pre-fill name and vendor on a scanned manufacturer barcode, but it is a bonus rather than a dependency. It never blocks, it falls straight to manual entry when it misses, and it is off until you bring your own lookup key in settings, so no shared key gets burned.

The storage map

When you want to know not just that you own something but exactly where it sits, the storage map gives you a location tree (freezer, then rack, then box, nested as deep as your storage really goes) with a box grid at the bottom. The grid is a rows-by-columns layout with A1-style positions, the same cell scheme a plate uses, and each filled cell shows the stock that lives there. Cells are colored by status, so a box of expiring or low tubes reads at a glance.

Click an empty cell to place a stock there, or a filled cell to see what it holds (item name, lot, expiry, count) and move or remove it. A breadcrumb above the grid shows the full path of the box you are looking at. The map is optional. You can run a perfectly useful inventory with nothing but the free-text location note, and reach for the freezer tree only when finding a specific tube is worth it.

Spreadsheet import for an existing inventory

If your lab already keeps a list somewhere (and most do, usually a spreadsheet), you do not have to re-enter it. The import path lets you paste straight from Excel or Google Sheets to bulk-load what you already have, so an existing inventory comes across in one action instead of one row at a time. Combined with the self-populating receive flow, this means the cold start, the part where most inventories never get off the ground, is close to free.

Item categories

Every inventory item belongs to exactly one category. The category you choose drives which extra typed fields appear in the detail editor. The full set of categories is listed below.

• Reagent (the default). Generic chemical or consumable with no typed extension fields.
• Antibody. Carries target, host species, clonality and clone, conjugate, isotype, reactivity, applications (for example WB, IF, IHC, FACS), recommended dilution, and an RRID for reproducibility. Antibodies are also the textbook case for opt-in precise consumption tracking, since they are expensive, finite, and often shared between users.
• Plasmid. Carries backbone, insert, resistance, bacterial host, size in base pairs, source and Addgene number, and an attached sequence file. Lines up with the cloning and sequencing work elsewhere in the app.
• Enzyme, Primer, Cell line, Strain. Plain categories with no extra typed fields today. A future version will add typed fields to each the same way antibodies and plasmids were extended.
• Kit. A multi-component commercial kit treated as a single inventory record. Track kit lots and expiry dates without splitting the kit into individual components.
• Equipment. A single piece of shared equipment. Equipment items have no container-count semantics since there is normally one instance, but they carry storage location, expiry (for calibration deadlines), and the same staleness signals as any other item.
• Other. For anything that does not fit the categories above.

All categories share the same stock list, storage positions, expiry, staleness signals, low-count threshold, history, trash, search, and sharing. The category only controls which extra fields the detail editor renders.

Precise consumption is opt-in, never the default

The volume-ledger model that sinks most inventories (deduct a few microliters per experiment) is the right tool for a narrow set of cases, like an expensive monoclonal antibody, a controlled substance, or a shared aliquot people argue over. For those, a precise per-use record is worth the effort because the stakes justify it.

So that behavior is an explicit per-item toggle, off by default. When off, which is the vast majority, the item lives entirely in the count, status, and expiry world above, and you never see a volume field you are obligated to maintain. When on, that one item exposes the deduct-from- stock workflow and a per-use audit trail. Inventory is fully useful with consumption tracking off for every item. It is a power feature you reach for when a specific reagent earns it, not a foundation everything sits on.