Assetize.Data WorkbenchProductUse casesCompareDocsOpen Workbench
Data Workbench · Docs

Prepare Maximo data with confidence

How the Assetize Data Workbench works, page by page, with what is built and what is planned. Every section says whether it is built today or planned.

Contents

Start here

What the Workbench doesQuick startWho it is for

The app, page by page

Dashboard and railEnvironmentsWhat are you loading?Start a new MaximoTry a sample loadVersion SetupSchema ExplorerIBM 9.2 referenceGrid EditorLayoutsKeyboard shortcutsLoad PlanPreviewMigration ReadinessRun ConsoleRun HistorySettings and Help

Tools and access

Dev ToolsPractice Lab and sample packsExports and the Excel editionSign-in and data handlingEnterprise self-hosting

Workbench 2.0

Workbench 2.0: the Engineering StudioFrom MXH finding to remediation evidenceThree levels of proofWorkflow in the StudioReports in the StudioDeveloper pack for EclipseNot built in the Studio yet

What the Workbench does

Built

Get Maximo data right before Maximo sees it. Bring a spreadsheet, fix it in a grid that knows Maximo, check the request, rehearse the load, retry only what failed.

Fix the sheet

Columns match the Maximo object you are loading. Red cells say what is wrong and what is allowed.

See the request

The URL, body and hash for the Maximo import, before anything is sent.

Rehearse the load

Run it against a simulator. Row-level results, a history, and retry for failures only.

Order the loads

Locations before assets, assets before PMs. The Load Plan sets the order and checks parents exist.

Check an upgrade

Compare a 7.6.1.3 source with a 9.x target and score what breaks before you load.

Pick your words

Optional. The account menu can show plain words, site and security detail, or Maximo's own names. The task is the same.

Good to know

Runs go to a simulator inside the Workbench, so you can rehearse a whole load with fictional or non-sensitive data.

Quick start

Built

From an open browser to a rehearsed load in six steps. No sign-in needed for the free web pilot.

  1. Open the Workbench at /workbench/. A guest workspace is created for you.
  2. On the Dashboard, drop your file or pick a saved load set. Or go to What are you loading? for an empty sheet or a sample, and to Start a new Maximo for the starter pack.
  3. Confirm the Maximo version in Version Setup. Choose Direct load, or Migration if you are moving between versions.
  4. Fix the sheet in the Grid Editor until the status bar says 0 errors.
  5. Open Preview to read the request. Nothing has been sent.
  6. Use Run Console to send the rehearsal. Read the results by row. Retry only failures from Run History.

Who it is for

Built

Ten kinds of people get value. Each has a full end to end journey on the journeys page. That journey opens the Workbench as its profile. Change it afterwards from the account menu.

Free

Evaluator

Decides in ten minutes whether this fits. Uses the fictional practice data.

Data loader

Data loader

Owns the spreadsheet. Wants it clean and loaded without calling a developer.

Maximo admin

Maximo admin

Cares about sites, organisations, security and what actually reaches Maximo.

Maximo developer

Maximo developer

Wants the exact payload, object structure names and the audit trail.

Migration

Upgrade lead

Moves 7.6.1.3 to MAS 9.x and needs to know what breaks first.

Delivery

Partner consultant

Delivers loads for customers repeatably, with packs and exports.

Viewer

Reviewer

Reviews the evidence or audits a load. Read-only access to runs and results.

Excel edition

Excel-first analyst

Lives in Excel. Uses the Excel add-in with the same checks.

Self-hosted

Platform owner

Keeps data inside the network. Runs the Workbench on their own server.

Studio

Upgrade engineer

Fixes the report or workflow an assessment found, and returns the evidence.

Dashboard and rail

Built

Home has one job: Load data into Maximo. Drop your sheet, try the sample load, or continue your last sheet; recent runs below, and upgrade and evidence work in one Other jobs row. The left rail has Home, Load, Schema, Environments, Studio and History; Settings and Help sit below. For a data loader it is just Load, History, Environments and Help.

Dashboard

A drop zone first: Drop an .xlsx or .csv, or pick a saved load set. The Workbench detects the object structure (an MXLoader workbook names it in its header rows), maps the columns and opens the grid. Try a sample load opens 18 fictional locations on the simulator with the Start here guide. Continue: <your sheet> opens where that sheet actually is (Fix 2 problems, Check, View results). Recent runs list each check and load with rows ok, failed and total. Other jobs: Assess an upgrade, Fix an upgrade finding, Evidence packs.

Project stepper

The steps of the job you chose, across the top. Load data: What are you loading, Grid Editor, Load Plan, Preview, Run Console. Migrate adds Version Setup, Schema Explorer and Migration Readiness before the load. Each step says where it is, in words as well as colour: done, needs attention, next, running or optional (the load order, for one sheet). That comes from your saved sheet, its checks, the load plan and the sheet's own run, never from having opened a page. An edit to the sheet, or a different environment, format or validation source, makes its check and run out of date, and so does editing the environment itself (its mode, or its address: host or path), even when it keeps its name and id. One rule decides this for the step bar, the grid's result strip and the Run Console. Each sheet keeps its own progress, so practice never takes yours away. Next names the real next action; on the Grid Editor the sheet's own button is the next action, so the header does not repeat it. Version Setup and Schema Explorer stay in the menu on a direct load. The header says where a check or run goes, Simulator or Live: host, at every screen width.

Start here guide

A row under the step bar while the sample load's Practice sheet is open: Clean the sheet, Check every row, Run it and read the result. It shows your progress, and Show me outlines the control to use: the first red cell, Check, Run simulation. It moves on only when the work is done. On your own sheets the sheet's one button and its Result column say what is next. Hide it with its X; Try a sample load or the command palette (Ctrl K, Open the Start here guide) brings it back.

Assetize logo

Top of the rail. Click it to go back to the main Assetize website. Home, just below, is the Workbench's own start page.

Environment pill

Shows which Maximo you are working against and its version. Click to switch.

Browse object structures

On each environment card. Makes that environment active and opens Schema Explorer on it: your Maximo's own object structures for a live connection, the simulator's otherwise.

Account menu

Your profile (Data loader, Maximo admin or Maximo developer), theme, resume link and sign-out. A journey opens the Workbench as its profile; change it here. Signing out clears this workspace's unsaved sheets, load plan and Studio projects from the browser, so the next person on it finds nothing. Saved load sets and runs stay on the server. Your other open tabs of the Workbench close their copy too, and a tab left open when someone else signs in reloads as their workspace. Remediation you returned to MXH from any of your tabs is removed from this browser; other people's stays.

Environments

Built

An environment is a Maximo target: a name, a version and a mode. Demo environments answer from the simulator.

ModeWhat happensStatus
DemoAnswered inside the Workbench. No request leaves it.Built

Requests come from this browser or Assetize cloud. This browser sends every Maximo request from your computer, so your network, VPN, internal DNS and address apply, and the key stays in this browser (sign-out removes it). The Workbench builds each request and reads the answer; it never reaches the host or keeps the key. Checks, loads, load plans, schemas and capture all work this way. Your Maximo admin allows this site once in System Properties: mxe.oslc.aclalloworigin = this site's address, mxe.oslc.aclallowheaders = apikey,maxauth,authorization,content-type,properties,x-method-override,patchtype, mxe.oslc.aclallowmethods = GET,POST,PATCH,DELETE,OPTIONS, then Live Refresh. The environment card has the lines ready to copy. Assetize cloud suits a public Maximo address and works from any device; the key is encrypted for your workspace.

Connect with the address and an API key or password. Saving tests the connection straight away, and the Workbench reads the system's fingerprint: the Manage version (so you never type it), every installed component with its build, the products and add-ons, the database, the app server, the base language and currency, and the default org and site. Personal details from the API user are not kept. The fingerprint shows on the environment card; Details lists every component.

Check egress IP on a live environment measures the Workbench server’s public address toward AWS checkip and shows its approximate IP-location city and region when available. The timestamp and probe route identify the observation. Requests use the same server fetch transport as Maximo; customer addresses and credentials are never sent to the diagnostic providers. Maximo can see a different source address: confirm it in the customer’s firewall or access logs. Cloud egress is not a fixed allowlist address, and an IP-location city does not select Sydney or Melbourne routing or establish data residency. The check is on demand, with no background polling; unavailable location data leaves the measured IP visible.

Development, Test or Production: read from the address or the name (dev, sandbox, test, uat, qa, sit, staging, preprod, prod, prd...). When neither says, the card asks once with three buttons, and your answer is kept through later edits. The kind shows on the environment switcher, the project bar and Review ("Goes to Production: host").

Export and import an environment: Export all in the page header, or Export on one card, saves an assetize-environment/1 file: the address, auth type, kind (dev, test, prod), version and fingerprint, every Object Structure schema this environment has read, with the customer's own structures and the fields outside the MAS 9.2 baseline flagged, and the objects and attributes behind them. Keys never go in the file. Import environments reads the same file from Workbench3 (assetize.com.au/workbench3), Studio or another workspace: each environment arrives as a live connection that runs from this browser, so open Edit to add its key. An address you already have is reused, not duplicated. Its Object Structures answer in Schema Explorer and checks straight away, before the key is entered, and a domain that came with the file puts its values on the fields that use it, so off-list values warn before a load. Once the key is in, the live list of Object Structures is read and the imported ones stay beside it.

Good to know

Credentials are never shown back. Headers in Preview are masked. Each account sees only its own environments.

What are you loading?

Built

Open the sheet you want in Maximo: drop your file or pick a saved load set, or start an empty sheet for an object structure (its key and required columns are set up). Sample sheets sit below. The Workbench opens the sheet already mapped.

Try a sample load always uses a simulator environment (one is created if you have none), never the environment you have selected, and opens a Practice sheet of its own. Your sheets, load plan and environment are left as they are; Back to my work returns to them. The Practice sheet stays on the simulator however you come back to it (its tab, a refresh, Back), and other environments cannot be picked while it is open. Opening it sends nothing: you still choose Check and Run, and confirm. Each start gives back the 18 sample locations: whatever was imported or typed into the Practice sheet since is replaced, and the guide starts again at step 1. If its simulator is deleted or edited into a live environment, Practice pauses: Check and Run stay off, the banner says why, and Use a simulator repairs it with a simulator that exists now. It never falls back to another environment. A workspace saved by an older version is repaired the same way.

Your file

Drop the .xlsx you already use with MXLoader, or a CSV. The object structure and action in its header rows are read as they are; otherwise the columns decide the structure, and you can change it above the sheet.

An empty sheet

Pick an object structure and Open empty sheet: the grid opens with its key and required columns, ready to type or paste into.

Sample sheets

Locations, assets, a messy asset sheet with six mistakes, PMs, a work order backlog, issuing stock to work orders, an inspection form and handover assets, each one click from the grid. Standing up a new Maximo? opens the starter pack on its own page.

Inventory usage: drop the flat MXINVUSE sheet Maximo exports (INVUSENUM, FROMSTORELOC, USETYPE, then line columns ITEMNUM, QUANTITY, REFWO). It opens as the structure the file is named after (MXINVUSE), or MXAPIINVUSE when the name does not say, with the line columns mapped to INVUSELINE, one row per line.

New numbers: put NEW (or NEW-1, NEW-2 to tell several apart) or leave INVUSENUM blank. Rows with the same placeholder become one usage, the placeholder is not sent, and your Maximo gives each usage its next INVUSENUM. A real number such as I_30002 is sent as typed. The same works for WONUM and ASSETNUM on parent and child sheets.

Issue the stock: put COMPLETE in STATUS. Maximo creates a usage at ENTERED and only issues it to the work order at COMPLETE, so Execute does both: it creates each usage (one request per usage, so Maximo returns the number it gave), then changes its status to COMPLETE. Rows left at ENTERED are created and stay there. Request and response lists every usage with its number and final status. Preview checks the whole sheet in one call and does not complete anything.

Accepted input: delimited text (comma, tab, semicolon, pipe, quoted values), pasted cells, JSON and XML. A new Maximo starts with the sets, the organization and the site. Practice locations are a separate sample. The simulator accepts any site code, so your own sheets preview with your own sites. When a sheet names an object structure the environment has no field list for (a customer structure such as WO_METERREADING_NEW on the simulator), Check and Run wait, the context bar's Checks says Pick an object structure, and a strip above the sheet offers Pick the object structure with the closest match listed first, plus Check on your live host when you have one connected. After an import the Workbench checks which object structure the columns fit; when another one maps clearly more of them it switches, and the sheet's structure and column mapping change together even if you have already moved on to Review, which then checks again.

Samples for assets, PMs and work orders carry a Golden check link. It opens that screen's golden script on assetize.com.au/ui/golden: drag the bookmark once, open the screen in your Maximo, click it, and see whether the title and controls match the MAS 9.2 screen harvest. Read-only. The free journeys kit runs every golden script at once (npm run golden).

Start a new Maximo

Built

Stand up a new Maximo from a starter pack at /packs. Pick the asset lifecycle modules — assets and locations, work management, inventory, preventive maintenance — and the add-ons and industry solutions on that Maximo. The sheets open in load order, organization and site first.

Use starter rows fills every sheet with one organization, ORG1, and one site, SITE1, carried through. Start with your own rows opens the same columns empty: import your file on each sheet. Each sheet can also be downloaded as CSV or opened on its own. Create a project from a template copies a whole project on the same page.

Try a sample load

Built

One sample load, at /practice, on the Dashboard and on What are you loading: 18 fictional locations at one site on the simulator, with the Start here guide. Check them, run them and read each row's result in the grid.

It opens its own Practice sheet. Starting it again always gives back the 18 locations, whatever was typed or imported into it since.

Version Setup

Built

Say which Maximo you are loading into. Checks use that version's rules, not a generic set.

Direct loadMigration
Versions askedTarget onlySource and target
ValidationTarget schema and version profileSame, plus migration-profile warnings
Schema Explorer diffHiddenShown
Migration ReadinessNot a step (still in the menu)A step before the load: scores the pair

Switch any time. The choice is stored with the project, saved load sets, runs and exports.

Schema Explorer

Built

Browse the object structures in your Maximo and their fields: type, length, which are required, which identify the record, allowed values. Open it from Schema in the left rail, or Browse object structures on an environment card.

Pick the source along the top: Your Maximo lists every object structure in the connected system, your own included, and opens by default on a live environment. Filter object structures above the list narrows it by name or description, with the count beside it.

In Migration, a second tab compares source with target: removed, added and changed fields. Data loaders see this as Browse fields.

Your fields puts your Maximo next to the IBM MAS 9.2 reference for the same object structure: the fields your team added (such as KE* or X* attributes) with their type, length and required flag, the IBM fields left out of your structure, the value lists you changed and your own child objects. Connect a live environment and it reads your system's schema; on the simulator it shows a labelled sample. Put your fields in the sheet as columns and the Grid Editor checks them like any other.

Your own object structures (such as KEASSET or XXWORKORDER) are listed with IBM's when you connect your Maximo. Pick one and load it like any other: its fields, types and lengths come from your system, and Check goes to your server, since no catalog holds it. Preview names the source ("KEASSET is your own object structure"), and Your fields shows every field in it as yours.

A flat export of your own multi-level structure (for example WO_METERREADING_NEW5310….csv: work order, multi-asset lines, asset, location, meters) opens as that structure, on the environment that has it (never a best-fit guess): the name comes from the file, Windows-1252 (Excel) files read correctly, and columns map to their level: first by the structure's own aliases, read from its definition in your Maximo (MAXINTOBJALIAS: ASSET-ASSETNUM is ASSETNUM on the asset), then by their prefix (WO-, ASSET-, LOC-, ASSETMETER-) or by where they sit, with a repeated name going to its own level. On load, rows fill down: the repeated work order becomes one record, each line, asset and meter attaches to the one above it, and the request goes as one nested JSON document per work order.

Good to know

Structure comes from observed lab metadata and version profiles. Where coverage is thin, the Workbench says so beside the result.

IBM 9.2 reference

Built

What IBM documents for the object and field you have selected, with the source named.

Reached from Help or the grid. Reference, observation and demo data are labelled separately. IBM statements are never mixed with Assetize interpretation.

Grid Editor

Built

The sheet is the hero. Type, paste from Excel, fix red cells, press one button, and read Maximo's answer on each row. Only the rows scroll: the toolbar, the notices and the column headers stay put, and 5,000 rows scroll as smoothly as 50.

Columns match Maximo

Mapped automatically from your headers. Unmapped columns are listed so nothing is silently dropped.

Cell checks

Type, length, required, allowed values and version rules. Hover or select a cell to read the reason in the formula bar.

Parent and child rows

Hierarchical objects such as inspection forms: parent cells blank below the first row. Rows that repeat the record key become one record with several lines: an inventory usage sheet (MXAPIINVUSE) with three rows for one INVUSENUM sends one usage with three INVUSELINE entries.

Problem list

Above the sheet: every problem with its row, column, value and what is expected, and Go to cell. Each says whether Fix problems handles it or it needs your decision. A required field that is not mapped ("Required field SITEID is not mapped") is a link: it takes you to the column that holds it (SITEID, WO-SITEID) and opens its field picker, or offers to add the column.

Fix problems

Rewrites formats only: hidden characters, dates to YYYY-MM-DD, Yes/No to 1/0, numbers without separators or currency signs. The button says how many problems it can fix; afterwards it says how many are left for you.

Result column

Frozen beside the row numbers, as in MXLoader: ✓ for a row that passed or loaded, or Maximo's message with its BMXAA code for one that failed. Editing a row clears its result. Rows with issues includes the rows Maximo refused. Each sheet keeps its own results, for the environment and format they ran on.

Run summary

One line above the sheet: "Check #12 · 17 ok · 1 failed", how many rows changed since, whether it was simulated, and Request and response. Retry failed sends only the rows whose last answer was a failure, fixed since or not, and their results update in place.

One button

Check until this exact sheet passes, then Load (Run simulation on the simulator). The other action is in the button's menu. A check and a simulator run start at once; a load into a live Maximo asks you to type its host first.

Get from Maximo

MXLoader's Query. Type a where clause in MXLoader syntax (siteid="BEDFORD" and status="OPERATING", status in ["A","B"], % as a wildcard) and an optional row limit. The Workbench reads the records with Maximo's export call (a GET with oslc.select, oslc.where and oslc.pageSize, up to 5,000 rows) into a new sheet named after the structure and the clause. Key columns are locked, the action is Update, and Check and Load send only the rows you changed. The simulator answers from its sample rows.

Action

Next to the object structure: AddChange (the default), Add, Update or Delete. An MxLoader workbook's own action is read from its header row. Anything other than AddChange is sent with each record as _action (Maximo's word: Add, Change, Delete), as JSON. Export writes the same action back.

Big sheets

Over 500 rows (Settings), Check and Load use Maximo's async import on their own and poll it; the footer counts the rows in, "Sending in the background · 1,240 of 5,000", and each row's result lands as it comes.

Before you load

One chip in the notices row opens the sites in this sheet, the parent order and, for admins, the loading account's checks. It stays closed until you open it and remembers your choice. The parent order shows on its own when a child sits above its parent or a row can never load.

Toolbar

Import, paste, open and save a load set, add rows and columns, match columns, Fix problems and Export. Export → MxLoader workbook (.xlsx) writes a RESULT column after the attributes with each row's last answer.

Layouts

Built

Pick the job and the Workbench sets itself up for it. Use the Layout menu in the top bar, or Alt+1 to Alt+5. The Workbench remembers your last layout.

LayoutForOpensWords and columns
PrepareData loaderGrid EditorPlain words. Only the columns that matter.
Map & checkMaximo adminGrid EditorAdmin words. Every column, with sites, organisations and rules.
UpgradeUpgrade leadMigration ReadinessAdmin words. Columns as you left them.
RunOperatorRun ConsoleYour current words.
ReviewApproverRun HistoryYour current words.

A layout is a shortcut for three choices: the page it opens, your profile, and whether the sheet hides optional columns. A journey does the same for its profile. You can still change each one yourself afterwards in Settings or on the grid. Dropping a file, opening /start or the sample load starts you in Prepare with the data loader's words, unless you picked a layout yourself; your choice of hidden columns stays yours. The data loader's rail is Load, History, Environments and Help, and the developer tools open only when you ask for them (Request and response, or Ctrl K).

Keyboard shortcuts

Built

The grid works like a spreadsheet. Press ? anywhere, or use More → Keyboard shortcuts, to see the list in the app. On Mac use ⌘ for Ctrl.

KeysWhat it does
← ↑ → ↓Move between cells. ← and → move when the caret is at the edge of the text.
Tab / Shift+TabNext or previous cell, wrapping to the next row.
Enter / Shift+EnterCell below or above.
Ctrl+← ↑ → ↓Jump to the edge of the filled block.
Home / EndFirst or last column in the row. Ctrl+Home and Ctrl+End go to the first or last cell.
PageUp / PageDownUp or down a screen of rows.
Shift+arrowsExtend the selection.
Ctrl+A, Ctrl+Space, Shift+SpaceSelect every cell, the column, the row.
Ctrl+Z / Ctrl+YUndo and redo. Ctrl+Shift+Z also redoes.
Ctrl+D, Ctrl+C, Ctrl+V, DeleteFill down, copy, paste (including from Excel), clear.
EscUndo the edit in a cell, then collapse the selection.
Ctrl+KJump to any page or Dev Tools tab. Arrow keys and Enter pick.
Alt+1 to Alt+5Switch layout: Prepare, Map & check, Upgrade, Run, Review.
Alt+Shift+← / →Previous or next step in the project.
? or Ctrl+/Show the shortcut list.

Load Plan

Built

Sets the order of loads and what each depends on, so children never arrive before their parents. Data loaders see this as Load order.

One step per sheet

Add the open sheet, every sheet with rows, or a parent-and-child sheet level by level. A sheet goes in one way only, whole or level by level: adding it the other way offers to replace its steps instead, so every row is sent once. Loading a sheet twice takes an explicit Load twice, with a warning that its rows are sent again.

Kept with your workspace

Steps, settings, the worked-out order and the last batch survive leaving the page, Back and refresh.

Edits are noticed

A step points at its sheet. Edit the sheet and the step says it changed, the plan uses the current rows, and the worked-out order is marked out of date until you calculate again. Queue stays off until it is current.

Rebuild when the levels change

If a parent-and-child sheet gains or loses a level, or a level step is removed, the plan says which sheet and offers Rebuild its steps. It cannot be calculated or queued until then. A removed sheet's step stays marked as removed; a new sheet never takes its place.

A batch keeps its own label

A queued batch says how it was queued (preview, or execute) and what its runs actually did: ran on the simulator, were sent to the host recorded on the run, or were blocked before anything was sent. It is read from the batch and its run records, so changing the Mode for the next batch, or editing the environment later, does not change it. Before a chunk has run it says where it runs is not known yet. Retries are listed too, each attempt by where it went and what its record confirms: checked by a host with nothing saved, written or partly written to a host, sent with no row confirmed written, still sending, run on the simulator, or blocked. The first attempt's sentence is labelled as the first attempt.

No accidental second run

Queueing the same plan again asks first. A failed batch offers Retry only the failed rows, as a preview; rows that succeeded are not resent.

Preview

Built

Review the request before it is sent: which sheet, how many rows, where it goes (Simulator, or Live: host) and what comes next. The URL, headers with secrets masked, body and hash are under Technical details, open by default for developers.

Check on simulator runs every row through the simulator, which saves nothing. Run simulation does the load there. On your own Maximo the buttons name the host. With the catalog as the validation source (the default), Check against catalog is answered here and sends nothing to your Maximo; with the server as the source, Check sends the rows to validate them without saving. Run writes them, and you confirm first. Each validation finding with a row opens the grid at it: Fix this cell when the check named the column, Review this row when it did not. Target-rule findings sit beside it. Data loaders see this as Check before loading.

Migration Readiness

Built

Pick object structures, run the assessment, get a score and the list of what breaks between source and target.

Optionally includes your grid's mapping. Data loaders see this as Upgrade check. Findings carry the evidence behind them, not just a colour.

Run Console

Built

Send the rehearsal and read the result: how many rows passed, which need fixing, and whether it was simulated (nothing written) or sent to your Maximo. Each row shows Maximo's own answer inline: the BMXAA code, then its words, next to the record's key (ASSETNUM · SITEID, LOCATION · SITEID and so on). Data loaders see this as Load and watch.

The header has one action, Fix failed rows in the grid: the run's sheet opens with each failed row's message in its Result column. Poll, Retry failed (as preview or for real), Clone as preview and Open as a new sheet are under More.

When Maximo refuses a whole request, every row carries Maximo's reason, read from either error shape it sends (Error with reasonCode and message, or oslc:Error with oslc:message), BMXAA code first. When the answer has neither, the row shows the HTTP status and the start of the body. More than five rows with the same message read as one line, "79 rows · BMXAA4195E · A value is required…", with Show rows to open them.

Data loaders see this as Load and watch. A run exercises the simulator's result paths: success, failure and retry. A card says what to do next, from the run's own record: after a passing check, Run; after failed rows, show them; after a finished simulated run, use your own spreadsheet or save a load set. A run counts as passed only from a success status with every row counted as passed; a failed, partial, blocked, running or unclear run never shows as done. The card's actions belong to the run's own sheet and target: when another sheet is open, it offers to open that sheet with the environment and format the run used, if that environment still has the mode and host the run went to; when the sheet or the environment has changed, or the environment is gone, it says so and offers a recovery copy. On the open sheet's own, unchanged run, each failed row has Review this row; once the sheet has changed, the page says the results are history and offers Open a recovery copy, which puts the run's rows in a new sheet. Open as a new sheet does the same; neither overwrites the sheet you have open. Clone as preview, Retry failed and a batch's retry go where the run went, or nowhere: the Worker checks before anything is sent. If the run's environment now points to another address, it says where the run went and where it would go now, and sends only when you choose Send to that address. A simulator run is never replayed to a live Maximo; open its rows as a new sheet and review the request instead. The Retry confirmation names where the retry would go now.

Run History

Built

Every run for your account: what was sent, what came back, by row. Retry failures only, or clone a run into a new one.

Runs, rows, audits and batches are stored per account. Viewers can read them but not change them.

Settings and Help

Built

Choose your profile, reset local state, find short explanations beside each page.

Data loader

Plain words. Only what the load needs.

Maximo admin

Sites, organisations, security and run settings too.

Maximo developer

Maximo's own names, API requests and payloads.

A journey opens the Workbench as its profile. Change it here or in the account menu. Reset clears the current project, grid and Dev Tools state in this browser. Server data is untouched.

Dev Tools

Built

A drawer for developers: the request and response, timings and the audit trail. Open it from the command bar.

The command bar (jump to any page or Dev Tools tab) is the quick way around the app.

Practice Lab and sample packs

Built

Ten curated templates from observed lab object structures, with fictional records. No real Maximo, no network send.

Use them to learn the flow, to train a team or to show a customer, without touching real data.

Exports and the Excel edition

Built

Take your work with you: export the sheet, the request or an MxLoader file. Or work in Excel with the add-in.

MxLoader workbook: sheet 1 is the object structure (row 1 names it with the action, row 4 the attribute headers, data from row 5, child columns as CHILD.FIELD with a blank parent cell continuing the record above), then a Mapping sheet and the Instructions. With an environment selected the Mapping sheet states each column's type, length and required flag from that environment's own dictionary, child columns included; without one, from the bundled catalog.

The Excel edition is a separate Windows product: an add-in (.xll) with a Getting Started workbook, sharing the same JSON contracts. It has its own ribbon for validate, preview, load, retry and history, and keeps customer data on your PC. It is free and needs no licence key. The shortcuts below are the web grid's. In Excel you have Excel's own. See /workbench/excel, and /workbench/compare for the web, Excel and self-hosted editions side by side.

Sign-in and data handling

Built

Free web pilot with a guest workspace. Sign in with a one-time email link to keep your work. Each account sees only its own data.

RoleCan
MemberCreate and edit environments, projects and runs
ViewerRead everything for the account, change nothing
Good to know

Bring fictional or non-sensitive data while the pilot runs. See /workbench/security for the data boundary.

Enterprise self-hosting

Built

Run the Workbench inside your own network. One Node 22 process and one SQLite file. Your data and saved credentials stay on a disk you own.

Sign-in you control

No guest workspaces. An administrator issues one-time links with node admin.mjs invite. Suspend and restore people the same way.

Host allow-list

You list the Maximo hosts the Workbench may be pointed at in MAXIMO_ALLOWED_HOSTS. Internal names and ports are fine. https is always required.

Sealed credentials

Saved Maximo credentials are sealed with your SESSION_SECRET and never returned by the API.

Backups and upgrades

admin.mjs backup copies the database while it runs. Migrations apply once at start.

Start with products/workbench-enterprise/README.md in the repository: Docker compose, plain Node, every setting and the admin commands.

Good to know

Single sign-on and high availability are not included. Back up the database file with admin.mjs backup.

Workbench 2.0: the Engineering Studio

Built

One project for upgrade work. MXH finds a problem, you open it here with its versions and evidence already loaded, fix it in the right editor, and send the evidence back.

Project

Source and target version, products and add-ons, the findings you opened from MXH, your artifacts and every proof they have earned.

Build

Workflow and Reports open as editors inside the project. Scripts runs the Script Test Harness in the project and on its own page: one scenario or the whole pack, including libraries, method results and related records. Data loads link to the Grid Editor. Your Maximo captures the scripts, launch points, workflow processes and applications your system has today.

Validate

Look up any object and attribute at the target version, see what your artifacts depend on, and see which proof each one has.

Ship

Download the project file or a developer pack for Eclipse. Every revision is kept: compare, restore, duplicate, export.

Open it from the rail, the Dashboard's Design Studio card, or an MXH finding with Open in Workbench 2.0. A Studio project is an upgrade project, separate from your data load; the Studio says so beside its versions. If the finding arrives before you have signed in, the Studio opens it once you have. The Evidence Graph is asked for facts and is never copied.

Data: add the sheet that is open in the Grid Editor to the project (up to 500 rows; a bigger sheet is refused with the reason, never cut short). UI: opens your sheet in a Maximo screen, or the Screen Designer on assetize.com.au.

Your Maximo: pick an environment and press Capture from my Maximo. The Workbench first reads that Maximo's MXINTOBJECT list (a server that does not list it, such as some 7.6 fix packs, is asked for each capture structure directly), then reads only the capture sets it has: automation scripts (MXAPIAUTOSCRIPT), launch points (MXAPILAUNCHPOINT), workflow processes (MXAPIWFPROCESS) and applications (MXAPIMAXAPP). Capture only sends GET requests, follows pages up to 500 records per set, and keeps the result in the project. Each capture also reads the version from the server, with the products and add-ons, so it never needs to be told which Maximo it came from; the panel says where that version sits in the project's upgrade. On the simulator it returns a labelled sample. Test runs and the UI page offer the same capture.

Environment files: beside the environment picker in Your Maximo, Export environment and Import environment move an assetize-environment/1 file between the Studio, the Data Workbench and Workbench3. An imported environment becomes the one the panel captures from; add its key under Environments, Edit.

Capacity: Size in the Capacity Lab opens the Capacity & Resilience Lab with this project's target version and the database, add-ons and kind of system (production or non-production) read from your Maximo's fingerprint. Set users, integrations, reports and mobile, then Save sizing to Studio: the project keeps worker nodes, cluster and Manage CPU and memory, pods per workload, database sizing, the model version and a link that reopens the scenario.

Be clear on this

Projects are stored in this browser, per workspace, not on a server. Signing out clears them. Export the project file to keep or share one. Projects an earlier version saved in this browser, before accounts, are not opened in any account: the Studio offers to download them as a file or discard them.

From MXH finding to remediation evidence

Built

The finding arrives with everything MXH knew. The evidence goes back with what actually ran. The finding stays open until its own closure test passes.

  1. In an MXH assessment, select a finding and press Open in Workbench 2.0. The source and target versions, object, attribute, owner, required action, closure test and evidence arrive with it.
  2. The Studio shows the context and looks the object and attribute up at the target version, with the evidence it rests on.
  3. Open Workflow or Reports on the object. The artifact is linked to the finding.
  4. Save and validate. Each artifact records its proof levels.
  5. On the finding, press Return evidence to MXH. MXH shows what was prepared and which proof levels passed.
Good to know

Returning evidence never closes a finding. MXH shows it as open, with badges such as Remediation prepared, Static validation passed and Target execution not run.

Three levels of proof

Built

Every result says what it is. The levels are never merged.

LevelMeansBuilt today
StaticThe artifact is structurally compatible with retained evidence.Yes: reports and workflows
SimulatedThe browser engine exercised our model of the artifact. Not Maximo runtime execution.Yes: workflow offline routes
Target executionThe artifact ran on that Maximo/MAS build and the result was retained.No. Always NOT RUN until a connected, authorised environment has run it
Be clear on this

A browser report preview and an exported .rptdesign are not BIRT execution. An offline workflow route passing is not Maximo running the process. Neither can produce a target-execution result.

Workflow in the Studio

Built

The Workflow Designer opens inside the project, in the shape of Maximo's own designer. It checks the fields your conditions name against the target version.

New builds a complete process from a pattern (approval, two-level approval, status cycle, request and fulfil, wait for an event, open a screen, or blank) on the business object you pick, with every route ready to prove. Open gives you the Maximo process samples, the industry demos, the lab WOSTATUS and import of designer JSON or a test pack.

The palette along the top holds the Maximo node types: Start, Task, Condition, Manual input, Interaction, Subprocess, Wait and Stop. Click one to add it after the selected step, or on its own. Drag from a step's port to draw a line: the right port is the positive line, the bottom port the negative one. Arrange lays the drawing out left to right with the main path along the top. Double-click a step for its IBM sheet.

Save a revision to run the designer's structure checks, your route scenarios and the version lookup. Conditions that run on a related object are listed as not checked, because the related object is not resolved here.

Reports in the Studio

Built

Build a BIRT report from a Maximo object, or import the .rptdesign you already have and see what changes at the target version.

Choose an object and columns, add parameters, filters, sort and grouping, preview on synthetic rows, validate against the target, and export a .rptdesign or a Power BI handoff.

Power BI: the Export tab downloads a zip with one Power Query (M) script per data set, a TMDL semantic model with typed columns, parameters and simple calculated columns, a report plan (page, tables, slicers) and a README listing what needs a person (event scripts, scripted expressions, libraries). BIRT layout and charts are not converted. The files are generated text; they have not been opened in Power BI Desktop here.

Import reads data sources, data sets and SQL, parameters, bound fields, scripts, libraries and layout counts. Every table and column is looked up at the target. Scripts that use Java or BIRT APIs are flagged for review.

Be clear on this

Retained MAS 9.2 evidence is Eclipse BIRT 4.21.0 with report scripts on Rhino 1.8.0. Other releases are not assumed to match. The importer reads Maximo-style scripted queries and was run on 338 Eclipse BIRT sample reports plus three Northgate example reports with no failures; the retained evidence holds no IBM .rptdesign files. Generated files have not been opened in the BIRT Designer here.

Developer pack for Eclipse

Built

For pixel-level BIRT layout, native libraries and Java work. The pack carries the same project file.

Download it under Packages. Open the reports in the native BIRT Designer, edit, and bring the .rptdesign back with Import. There is no second Workbench in Eclipse.

Be clear on this

The pack is generated. It has not been opened in Eclipse by this tool. An Eclipse plug-in and Java extension tooling are planned.

Not built in the Studio yet

Planned

These are named here so nobody has to guess.

ItemState
UI and Integrations editorsPlanned. They link to the UI Playground and the Integration Flow Explorer.
Target execution runner for workflows and reportsPlanned. Today the Studio checks artifacts against the evidence; connecting an environment does not run them.
Projects stored on the serverPlanned. Today a project lives in this browser.
Excel add-in button that opens a Studio linkPlanned. The links exist; the add-in does not carry them yet.
Nothing matches that search.