For many hydraulic modelling teams, the bottleneck is not skill or ambition; it is hardware. In the latest instalment of the Autodesk Water webinar series, Hydraulic Modelling in the Cloud with InfoWorks ICM & InfoWorks WS Pro: Lessons from Scottish Water's Early Adoption Experience, held on 4 June 2026, Samer Muhandes, Senior Product Manager for InfoWorks ICM at Autodesk Water, and Stephen Leslie, Hydraulic Principal Digital Specialist at Scottish Water, made the case that cloud-based workflows offer a practical way out of that constraint, drawing on Scottish Water's own experience as an early adopter.
The discussion centred on a question familiar to modelling teams everywhere: what becomes possible when simulation capacity is no longer limited by local hardware? Climate change, tighter environmental standards, and increasingly complex networks are pushing traditional workflows to their limits, and utilities are exploring how to work at greater scale without compromising speed or collaboration. Long simulation runtimes, fragmented datasets, version control challenges, and the need to run multiple scenarios simultaneously have become routine pain points, and the session set out to examine how cloud adoption can address them.
From the office weekend to the cloud
It’s not only a technology shift — it’s improving the collaboration, the accessibility, and you’re now less dependent on being in the right office at the right desk at the right time - Samer Muhandes
Muhandes opened with a personal story that set the tone for the session. Facing a project deadline with simulations that had to run over a weekend, he found himself taking his young daughter into the office because the only machine powerful enough was locked to a physical workstation. The point was not nostalgia: it was a diagnosis. When a modelling workflow depends on being in the right place at the right desk, competing with colleagues for shared hardware, it becomes fragile. Anyone who has worked in a consultancy modelling team will recognise the situation: negotiating access to a virtual machine, queuing simulations behind a colleague’s project, or finding that the hardware required for a large catchment is simply unavailable at the moment it is most needed.
The InfoWorks cloud offering, available for both ICM and its water distribution counterpart InfoWorks WS Pro, is designed to remove that fragility. Engineers can store model data and simulation results centrally, run up to 80 simulations in parallel on high-performance hardware, and work from any location without needing to plan for local capacity. Whether a simulation is 1D or 2D, the platform allocates the appropriate computing resource automatically. The result, Muhandes argued, is not just a faster machine; it is a fundamentally different relationship between engineers and the questions they can afford to ask. Scenarios that were once computationally prohibitive become routine.
A key part of the presentation covered how organisations can structure access to cloud databases. Projects sit within regional hubs, and access is governed by roles: editors can modify models and run simulations, while readers can view results, plot hydrographs, and export outputs. Crucially, readers do not need a paid licence. A free ICM Viewer, bundled with the standard installation, allows clients, reviewers, or regulatory bodies to access shared cloud databases without a subscription. This changes the economics of model sharing considerably, enabling broad access to outputs while retaining full editorial control. A utility can create a dedicated project for a consultant, grant editor access, and add an external reviewer as a reader — all without exposing any other part of their model library.
Scottish Water’s migration journey
Stephen has been at Scottish Water for 18 years and was part of the team that rebuilt the hydraulic modelling function in-house 13 years ago, starting from a handful of standalone machines with no consistent version control. Work was getting lost, model updates were not reaching everyone, and the team was working in isolation. Over the years that followed, the team grew to around 33 modellers covering both water and wastewater, managing 330 wastewater models and 120 water distribution models, with approximately 70 to 80 viewer users.
Due to the cloud, we have been able to bring this in-house for the 65 catchments. Previously, every catchment was taking approximately 35 days. We are now able to turn out a Section 16 within approximately 10 days - Stephen Leslie
The water side has now fully transitioned to the cloud over the past 18 months. The wastewater side operates on a hybrid basis: cloud is used project-specifically, selected where computational demands justify it, while on-premises systems handle the approximately 350 transportable models issued to external consultants each year. Stephen was candid about the process — gaining internal approval in a large organisation required building a business case over several years, engaging digital teams, and clearing security and data governance requirements. Autodesk provided documentation confirming that data would be stored within Europe before Scottish Water gave the green light.
Stephen Leslie shared benchmarking data from a 1D model of the Aberdeenshire and Aberdeen area, run identically on both the on-premises system and the cloud. On-premises, a single simulation took 120 to 130 minutes. On the cloud, the same simulation completed in 17 to 20 minutes — an approximately 80 per cent reduction. Both speakers were careful to contextualise the figure: it reflects Scottish Water’s specific infrastructure, and results will vary depending on the hardware organisations have invested in on-premises. But the direction is clear, and Stephen noted that where his on-premises system could handle 35 to 40 parallel simulations, the cloud supports 80 at significantly faster speeds.
Two projects illustrated what this means in practice. A climate change investigation required 15 models to be run across 40 years of rainfall data. Done on-premises, the computational load would have displaced all other business-as-usual output across the whole of Scotland. With cloud capacity, the work was completed in 12 weeks. A Section 16 investigation currently underway covers 65 catchments with full 2D meshes. Combined with Ruby script automation developed by Stephen’s team, turnaround per catchment has fallen from around 35 days to approximately 10.
Practical advice and Q&A
The Q&A drew questions from organisations including Jacobs, Southern Water, Severn Trent, Arup, and Veolia. Several themes recurred. On data architecture, Muhandes clarified that the preferred setup is ICM installed locally, with databases hosted on Autodesk’s AWS environment; running ICM from within an Azure virtual machine introduces dual gateway traffic and potential latency. On vendor dependency, he was explicit: migration is not a one-way decision. Data remains accessible at all times, models can be moved back on-premises whenever needed, and the on-premises offering is not being retired. He described it as a toll road: the original route still exists, but the new one gets you there faster. You choose which road to take.
Stephen Leslie added two practical points from experience. On security: Scottish Water’s approval process was thorough but navigable. Autodesk responded to every documentation request and confirmed data residency requirements in writing. On local drive management: once results have been analysed, delete local copies. They remain stored on the cloud and can be re-downloaded at any time, but unchecked downloads accumulate quickly and can fill a local drive without warning.
His broader advice to utilities considering the transition was to build incrementally, design model tree structures with viewer users in mind, not just modellers, and resist the temptation to migrate everything at once. Start with what the business needs today, and build toward what it will need in the future.
Both speakers closed on the same note. Cloud adoption is not a technology decision in isolation; it is a decision about how a modelling team spends its time and what it is able to deliver. Less managing infrastructure, more answering the questions that matter. For Scottish Water, the journey was gradual and deliberate, and the benefits are now tangible: faster simulations, larger investigations brought in-house, and results shared directly with collaborators through project hubs. For utilities considering a similar path, the message was straightforward: start where you are, build to what you need, and try it.
