DAARWIN
Cloud platform from Spanish firm SAALG Geomechanics that centralises geotechnical data across a project's life cycle, combining ground investigation records, live monitoring and finite element models to calibrate soil parameters automatically for construction and mining works.
What is DAARWIN?
DAARWIN is a cloud platform for geotechnical engineering, developed and sold by SAALG Geomechanics SL, a Barcelona company that grew out of the geotechnical division of the Universitat Politecnica de Catalunya BarcelonaTech. Its stated claim is to be the first and only cloud platform that connects and centralises all geotechnical data across an entire project life cycle, for the construction and mining industries.
The platform is organised as seven modules around a shared data repository. The Public Ground Information System aggregates publicly available ground data, more than six million ground information points across twelve countries, and is free to use. Ground Investigation Data Management turns raw survey records into a full digital ground model. Instrumentation & Monitoring centralises live readings from inclinometers, piezometers and other sensors so that trends and anomalies surface early. TBM Data Management pairs tunnel boring machine performance with predictive modelling to detect ground anomalies as excavation advances.
Two modules carry the heaviest computation. Sensitivity Analysis measures how much each parameter influences a design. Real-Time Back-Analysis is the centrepiece: it couples monitoring data with finite element models and uses genetic algorithms to adjust ground and structural parameters iteratively until predictions match what the instruments actually record. Both run on a parallel cloud architecture the vendor describes as up to two hundred server instances across four hardware tiers, which is what turns weeks of manual fitting into repeatable runs. PLAXIS models import directly, with all constitutive models supported.
The seventh module, Digitisation Tools, is where the platform is most visibly an AI product. It combines optical character recognition, AI-based extraction and machine learning, including image classification, object detection and text interpretation, to convert scanned boring logs, laboratory results and full geotechnical reports into structured data, handling low-resolution scans and handwritten notes. The vendor states that reports of several hundred pages are processed in under an hour.
The intended frame of use is the Observational Method, where design is revised as construction reveals the ground. Named users include ACCIONA, Ferrovial, Deutsche Bahn, Transport for London, Mott MacDonald, COWI and Kajima.
What it does
- Calibrate soil and structural parameters automatically against live monitoring data
- Run sensitivity analyses on finite element models using parallel cloud computing
- Build a full digital ground model from investigation data
- Digitise legacy geotechnical reports and boring logs through OCR and AI extraction
- Centralise instrumentation and monitoring readings and detect anomalies
- Track tunnel boring machine performance against predictive models
- Search aggregated public ground information across twelve countries
When to use DAARWIN / When not to
A quick filter to help you decide if DAARWIN is the right fit.
When to use DAARWIN
- Geotechnical engineers calibrating finite element models against real monitoring data
- Tunnelling teams running TBM works who need ground anomalies flagged as excavation advances
- Contractors and consultancies applying the Observational Method on deep excavations
- Mining operators monitoring slope, tailings and underground stability
- Organisations holding decades of paper geotechnical reports they want digitised and searchable
When not to use DAARWIN
- Buyers who need a published price before starting a conversation, as every plan is quoted individually
- Short projects unable to accept the twelve-month minimum subscription commitment
- Teams that need to drive the platform from their own software, as no public API exists
- Field staff expecting a mobile application, which the vendor publishes on no store
- Engineers working outside geotechnics, as nothing in the product addresses other disciplines
How to use DAARWIN
A typical end-to-end flow, from setup to results.
- Open the free Public Ground Information System portal and check what public ground data already exists for the site
- Request a quote from SAALG, since access to the full platform is sold on quotation only
- Agree the plan, the number of users and the storage volume, and settle any initial set-up fee
- Upload existing ground investigation records to build the digital ground model
- Send historical paper reports and boring logs through the digitisation module to bring legacy data in
- Import the project's PLAXIS finite element models
- Connect the instrumentation and monitoring feeds, read manually on the entry plan or automatically on the custom one
- Run a sensitivity analysis to identify which parameters actually drive the design
- Launch back-analysis so the platform calibrates parameters against the measured response
- Compare back-analysed results, monitoring data and model outputs in a single view, and feed the conclusions back into the design
Pros & Cons
Pros
- Covers the whole geotechnical life cycle in one platform rather than a single stage
- Explicit PLAXIS compatibility, with all constitutive models supported
- Substantial parallel computing capacity announced, up to two hundred server instances
- A genuinely free module giving access to more than six million public ground data points
- Named, verifiable industrial references including ACCIONA, Ferrovial, Deutsche Bahn and Transport for London
- Backed by academic research at UPC BarcelonaTech, with two professors on the technical advisory board
- Digitisation module handles degraded material such as low-resolution scans and handwritten notes
Cons
- No price is published anywhere, so no entry ticket can be assessed before contacting sales
- Every subscription carries a minimum commitment of twelve months, with a possible set-up fee
- No Terms and Conditions are published, although the privacy policy explicitly refers to them
- No API and no public technical documentation, so the platform cannot be driven from other software
- No free trial is offered or mentioned
- The privacy policy dates from March 2019 and mentions a mobile application that does not exist
- Published contact details are inconsistent, with displayed phone numbers differing from the link behind them
Pricing & Plans
There is no published price. SAALG Geomechanics presents tailor-made plans and invites visitors to request a quote, so no entry point can be stated in any currency. The structure is nevertheless described: a monthly subscription for platform access, usage-based charges for sensitivity and back-analysis runs, digitisation sold as a service, and an additional monthly subscription for the TBM module. Every subscription carries a minimum commitment of twelve months, after which it may be terminated with fifteen days' notice before the end of the month, and an initial set-up fee may apply depending on the client's requirements. One component is free of charge: the Public Ground Information System.
- five users
- 250 GB of storage
- manual instrumentation data acquisition
- sensitivity and back-analysis functionality included
- digitisation offered as a service
- TBM functionality not available
- price on quotation
- user count and storage capacity set ad hoc
- manual and automatic data acquisition
- sensitivity and back-analysis functionality included
- digitisation offered as a service
- TBM functionality available
- price on quotation
- free of charge
Data, GDPR & hosting
A consolidated view of how DAARWIN handles your data.
GDPR overview
Implementation is documented, but compliance is never claimed. The privacy policy, effective 8 March 2019, carries two dedicated GDPR sections. It names four legal bases: performance of a contract, consent, legitimate interest and legal obligation. It enumerates the rights of access, rectification and erasure, objection, restriction, portability and withdrawal of consent, and points users to their national supervisory authority in the European Economic Area. Personal data is transferred to Spain and processed there. The publisher is established in Barcelona, so no Article 27 representative applies and none is named. What the site never does is state in so many words that it is GDPR compliant, which is why that flag is left unset rather than asserted.
Who owns the data?
The published documents do not settle this. SAALG Geomechanics SL presents itself as the data controller for personal data collected through the website, and lists what it gathers: name, email address, postal details, cookies and usage data. Nothing in the privacy policy addresses ownership of the geotechnical project data that customers upload, which is the material that matters most for this tool. The policy refers to a set of Terms and Conditions that would normally carry those clauses, but no such document is published at any address on the site. Ownership of project data therefore remains publicly undocumented.
Reuse rights
Reuse rights are not documented. The privacy policy explains what the publisher does with personal data, namely delivering and maintaining the service, providing customer support, producing analytics to improve it, detecting incidents and sending news about similar offerings, with an unsubscribe link for marketing messages. It says nothing about whether customers may reuse, export or redistribute the ground models, digitised reports and back-analysis results produced in the platform. The Terms and Conditions the policy points to are not published anywhere on the site, so no permission and no restriction on reuse can be quoted from the vendor.
Data retention & training
Hosting summary
The vendor says little, and what it says covers the website rather than the platform. The privacy policy states that data, including personal data, is transferred to Spain and processed there, which places the jurisdiction inside the European Union. No hosting provider is named, no data centre or cloud region is identified, and no hosting certification is claimed anywhere: a search of every collected page returns no ISO 27001, no SOC 2 and no equivalent. Technically, the marketing site and the product sit on different infrastructure. The public pages are served from a website builder on an anycast address that reveals nothing about where the company operates, while the DAARWIN application answers on a separate host. That separation matters, because the hosting of the showcase site says nothing about where customer ground models, monitoring series and digitised reports actually live. Anyone with data residency obligations should ask the vendor directly, since the published material does not answer the question.
Things to keep in mind
Risks and trade-offs to weigh before adopting DAARWIN.
- Automated calibration can breed false confidence: genetic algorithms always return a best-fit parameter set, even when the monitoring data feeding them is sparse, drifting or from a badly installed instrument
- Deskilling is a real hazard on safety-critical works, because if back-analysis becomes a button rather than a judgement, engineers may lose the feel for when a ground model is simply wrong
- Ownership of uploaded project data is undocumented, as the privacy policy covers personal data only and the Terms and Conditions it refers to are published nowhere
- The privacy policy dates from March 2019 and has never been revised, which is old for a platform that has since added AI-based modules
- The publisher states that its policy covers a mobile application, yet no such application exists on any store, a sign the legal text has drifted from the product
- The contact block contradicts itself, displaying two phone numbers while its own link points to a third
- No price is public and a twelve-month minimum commitment applies, so the real cost of ownership cannot be assessed before committing to a sales process
Setup & Integrations
Technical difficulty
Moderate, and more demanding in expertise than in installation. Access is through a web browser with nothing to install, and the free public ground data portal opens straight away. The effort lies elsewhere: the platform expects existing PLAXIS finite element models and the skills to interpret them, plus instrumentation feeds to connect, read manually on the entry plan or automatically on the custom one. An initial set-up fee may apply depending on requirements, which suggests configuration is done with the vendor rather than alone. No public documentation or onboarding guide is available.
Deployment
Integrations
Supported languages
Behind DAARWIN
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
Who publishes DAARWIN?
How much does DAARWIN cost?
Is any part of DAARWIN free?
Which countries does the public ground data cover?
Does DAARWIN offer an API?
Is there a mobile application?
Does DAARWIN work with PLAXIS?
What technology drives the back-analysis?
Where is the data processed?
Is there a minimum age to use the service?
Should you pick DAARWIN?
DAARWIN is a credible, narrowly focused tool rather than a general-purpose AI product, and it is strongest where geotechnical practice is weakest: turning measurements taken during construction back into a corrected ground model. The chain it assembles is coherent from end to end, from free public ground data, through investigation records and digitised legacy reports, to instrumentation feeds and automatic parameter calibration. The AI content is real and specific rather than decorative: genetic algorithms for calibration, and optical character recognition with machine learning for document digitisation. The industrial references are named and checkable, and the academic lineage at UPC BarcelonaTech is consistent with what the product does.
The reservations are commercial and documentary rather than technical. Nothing about the price is public, the minimum commitment runs to twelve months, and a set-up fee may apply, so a prospective buyer cannot size the investment without entering a sales conversation. More awkwardly for a vendor handling client project data, the site publishes no Terms and Conditions at all, even though its own privacy policy refers to them, and that policy has not been revised since March 2019. It still describes a mobile application that exists on no store. Ownership and reuse of the ground models and digitised reports customers put into the platform are therefore undocumented, which is the single point most worth raising before signing.
The company itself is active and long-established for a startup, incorporated in 2016, with a registered office transfer recorded in May 2026 that the website has not yet reflected. For a geotechnical team already working with PLAXIS models and instrumented sites, DAARWIN is worth a conversation; the diligence should focus on the contract, not on the engineering.
- Choosing a selection results in a full page refresh.
- Opens in a new window.