Power BI

Power BI is the most common way Ditio customers build their own dashboards — hours per project, machine use, absence, HSE follow-up. Ditio’s Power BI template signs in with your API client and loads your data through the Data Extraction API, so you can go from credentials to your first report without writing queries.

Download

Download the Power BI template (Ditio-PowerBI-mal.pbit)

• The report is in Norwegian. Fields have the same names as in Ditio’s Excel export (“Kostpris maskin”, “Stilling”, “Prosjektnr” …).
• User guide (Norwegian): setup, publishing and troubleshooting, step by step.
• Field glossary: every field (report name, API field, Excel column) and every measure.

The template contains no data, credentials or tokens. You enter them when you open it. The template version is shown on the report’s start page.

What’s in the template

Table	Contents	Endpoint
Timeføringer	Hours for people, machines and vehicles, with project, work order, approval, payroll approval, cost and price	v1/time-registrations
Maskinregistreringer	Machine hours per machine, driver and project	v1/machine-registrations
Fravær	Absence per person and day, with absence type and approval	v1/absence-registrations
Varsler	HSE, quality, environment and machine alerts with status	v1/incident-registrations
Sjekklister	Checklists and forms with status, template and deviations	v1/checklist-registrations
Massetransport	Trips with quantity, mass type, load and dump area, cycle time and distance	v1/flow-trip-registrations
Varetransaksjoner	Item consumption, sales, purchases and adjustments	v1/item-registrations/web-query
Prosjekter, Arbeidsordrer, Ressurser, Brukere	Lookup tables to filter and group by	v1/project, v1/project/work-breakdown-structure, v1/resource, v1/user
Dato	Calendar with ISO weeks (Monday first), months and years	—

The template also includes:

• 45 ready-made measures in the Målinger table, grouped into folders:
  • hours (person, machine and vehicle hours; approved, payroll-approved, and unapproved for more than 14 days);
  • machines;
  • absence;
  • HSE and quality (alerts per 100,000 person hours, checklists with deviations);
  • mass transport (tonnes and m³ are never added together);
  • items;
  • economy (cost, sales, contribution margin, and how complete prices are);
  • data status (rows and newest change per table, used by the start page).
• A relational model. Every registration table is linked to the date table, and all except item transactions to projects. Hours are also linked to work orders, resources and users; absence to users; machine registrations to resources; alerts and mass transport to work orders. A date slicer filters every table; a project slicer filters every table except item transactions.
• A start page that shows the loaded period, the last refresh, and the rows and newest change in each table, so you can see at a glance that the data is complete and fresh.
• Brukere has one row per person. People with profiles in several companies (project companies, subsidiaries) are shown with the profile from their employer company.

Prerequisites

1. Power BI Desktop (Windows).
2. An API client. An administrator creates it in Ditio Web.
   • Create a dedicated client for Power BI with only the reportingapiv1 scope. Don’t reuse a client that also has ditioapiv3.
   • The user it acts as must be an administrator to read projects, resources and users. Because of that, everyone who can view the published report sees all loaded data, including absence and cost prices, regardless of their own access in Ditio. Limit who can view the report, or add row-level security in Power BI.

Set up

When you open the template, Power BI asks for these parameters. Values in parentheses are for the test environment; use an API client created in the same environment.

Parameter	Value
ReportingApiUrl	https://integration.ditio.no/reporting (test: https://core-api.ditio.dev/reporting)
IdentityUrl	https://identity.ditio.app (test: https://identity.ditio.dev). Must be the same environment as ReportingApiUrl.
CoreApiUrl	https://integration.ditio.no (test: https://core-api.ditio.dev/core). Only used by queries you add against the Integration API.
ClientId / ClientSecret	Your API client. The template requests a fresh token on every refresh, as long as AccessToken is empty.
AccessToken	Leave empty. If set, it overrides ClientId/ClientSecret and is never renewed, so scheduled refresh fails once it expires.
RangeStart / RangeEnd	The period loaded in Power BI Desktop, for example 01.04.2026 00:00 to 01.07.2026 00:00 (RangeStart included, RangeEnd not). A few months is enough if you publish: the Power BI service then loads the last 24 months itself — see How data stays current. If you only use Power BI Desktop, set the whole period you want.
CompanyId	Optional: only load data registered in one company of your group. Leave blank if unsure.

When Power BI asks how to connect to the two Ditio addresses, choose Anonymous: the template signs in by itself. Set the privacy level of both to Organizational.

Every endpoint except item transactions is read page by page, so long periods work. Each table gets a fresh token, so it must finish loading within the token lifetime; if it doesn’t, the refresh fails with HTTP 401 — shorten the period.

How data stays current

The registration tables (hours, machine registrations, absence, alerts, checklists, mass transport and items) use Power BI incremental refresh:

• In Power BI Desktop only RangeStart–RangeEnd is loaded.
• In the Power BI service each table is split into monthly partitions. The first refresh after publishing loads the last 24 months — it takes the longest. Every later refresh re-reads the current month and the two months before it; older months are kept as they are.
• Projects, work orders, resources and users are always reloaded in full.

Each partition is fetched with its own date range and paged with the API’s continuationToken, so a refresh downloads about three months of registrations instead of the whole history. Alerts and checklists are read as “changed since the partition start” and then limited to the partition.

Older registrations are frozen

Registrations dated before the first day of the month two months back are not re-read. On 15 October, that is everything dated before 1 August. A late approval, payroll lock, correction or deletion on such a registration doesn’t reach the report until it is reloaded in full: publish the report again from Power BI Desktop.

• Changing the policy: to keep more months or re-read more, right-click the table in Power BI Desktop → Incremental refresh, before publishing.
• Keep your .pbix file. A model with incremental refresh can’t be downloaded back from the Power BI service. Until the first refresh after a republish has finished, the report shows only the Desktop period.
• First refresh too slow? Power BI Pro stops a refresh after 2 hours. If the first 24-month load times out, lower the number of stored months before publishing.

Need a true change feed?

Power BI can’t keep a sync position between refreshes, so it can’t apply the Data Extraction API’s change feed (ModifiedSince, including deletions) itself. To keep a database exactly in sync and point Power BI at it, see the incremental sync example.

Scheduled refresh

After publishing to the Power BI service, open the semantic model’s Settings → Data source credentials. For both the Reporting API and the identity server:

• Authentication method: Anonymous
• Privacy level: Organizational

Caution

Check “Skip test connection” for both data sources. Power BI’s test connection is sent without the template’s token and always fails — the actual refresh works fine. No gateway is needed.

Treat the report file as a credential

ClientSecret (and AccessToken, if used) is stored in plain text in the .pbix and in the published semantic model. Anyone who can open, download or edit the report can read it and call the Data Extraction API as an administrator, outside the report too.

• Use the dedicated reportingapiv1-only client from Prerequisites.
• Share the file and workspace edit rights only with people who may see all of the company’s data.
• If the file leaks or someone with access leaves, rotate the secret or disable the client (see Creating API credentials).

Good to know

• Alerts per 100,000 person hours is a reporting rate, not H1/H2. Ditio doesn’t register lost-time injuries.
• Economy measures are only as complete as the prices registered in Ditio. The measure Andel timer med kost (share of hours with a cost price) shows how complete they are.
• Dates are shown as the API returns them, without time-zone conversion.
• Item transactions have no ids, only names and project numbers, and a double quote (") in an item name or description currently breaks parsing of that row.

Upgrading from the v1 template

Templates before v2 (Ditio api - data source examples v1.x.pbit) read Ditio endpoints that are not part of the public API and can change or stop working without notice. Move your reports to the new template. The README maps every v1 table to its replacement. Each query is also available as a separate file you can paste into an existing report.

Common issues

Issue	Cause	Fix
”Ditio-pålogging mangler”	Neither ClientId/ClientSecret nor AccessToken is filled in	Fill in ClientId and ClientSecret under Transform data → Edit parameters
”Ditio-pålogging feilet” with invalid_client	Wrong ClientId/ClientSecret, or credentials from the other environment	Check the values and that IdentityUrl matches the environment the client was created in
”Ditio-pålogging feilet” with invalid_scope	The API client lacks the reportingapiv1 scope	Contact [email protected]
”Ditio API-kall feilet” with HTTP 401	IdentityUrl and ReportingApiUrl point to different environments; a pasted AccessToken has expired or lacks reportingapiv1; or one table took longer than the token’s lifetime to load	Use matching production (or test) values for both URLs; clear AccessToken and use ClientId/ClientSecret; shorten the period. See Error Handling
”Ditio API-kall feilet” with HTTP 403	The API client’s user isn’t an administrator	Ask an administrator to check the API client in Ditio
Formula.Firewall error	Privacy levels not set, or set to Private	Set both Ditio addresses to Organizational (File → Options and settings → Data source settings)
“Test connection failed” in the Power BI service	The test is sent without the token	Check Skip test connection
A table has 0 rows on the start page	The company doesn’t use that part of Ditio, or the period is wrong	Check RangeStart/RangeEnd and CompanyId

Related

• Power Query Examples — the template’s queries, to reuse in your own reports
• Data Extraction API — every endpoint and field
• Authentication — creating API credentials
• Export Time Data — the API-first alternative for data warehouses