with unicorns like Facebook, Netflix or Uber. However, the sector needs to avoid
being a slow adopter of data science and consequently should consider investing
decisively in this direction. Solutions include:
• Some key recommendations to drive the digital agenda more generally are as
follows:
– First, secure executive buy-in, and then devise a digital strategy with an action
plan (and stick to it).
– Second, build the technological foundation by ensuring the basics are in place
to support future growth.
– Third, focus on business priorities, and communicate quick wins to tie the
investment in digital to outcomes that support the strategy (immediate pay
back is often needed in order to gain approval for implementation).
• Promote in-house expertise and roles (e.g. data scientist) hence embedding
personnel within water companies. Leveraging existing toolboxes (particularly
for machine learning) and workflows is becoming easier so that the barrier to
entry has now been now lowered (PhD qualifications were usually required in
the past for AI).
• Adopting open-source programming languages (such as Python) and tools
can facilitate collaboration and shared development. Software produced in
water engineering research has previously often been bespoke and stand
alone, and more thought needs to be directed at reusability, sustainability and
maintainability. For example, university research project software is usually
developed in isolation in languages such as MATLAB and may only persist
through one or two generations of researchers (such as passed down from
an academic to a PhD student).
• In the UK, centres for doctoral training like STREAM and WISE embed a
doctoral engineer in water companies helping such transfer.
• Other collaborations between industry, SMEs and research establishments in
projects (e.g. through European H2020 or UK Innovate funding) encourage
knowledge transfer. Data dives and hackathons provide forums for building
teams of collaborators.
Some private sector water utilities are already leaders in the digital space
and active in sharing their early successes with leveraging connected devices,
IoT and machine learning. This digital re-imagination of the sector will enable
a broad spectrum of outcomes from improved efficiencies to optimised asset
management across providers and potentially new business models such as
consolidated multiutility retail operations. While there is an increase of digital
adoption in water, the sector still lags behind other industries in integrating
new, smart technologies. Water utilities can benefit from the lessons learned in
other sectors (such as energy) and the established best practices and network
infrastructures. As technology evolves, the price for smart devices decreases,
the functionality increases, and consequently piggybacking on other sectors’ lead
can result in an accelerated adoption rate for realisation of benefits in the water
ecosystem.
22
S. R. Mounce
being a slow adopter of data science and consequently should consider investing
decisively in this direction. Solutions include:
• Some key recommendations to drive the digital agenda more generally are as
follows:
– First, secure executive buy-in, and then devise a digital strategy with an action
plan (and stick to it).
– Second, build the technological foundation by ensuring the basics are in place
to support future growth.
– Third, focus on business priorities, and communicate quick wins to tie the
investment in digital to outcomes that support the strategy (immediate pay
back is often needed in order to gain approval for implementation).
• Promote in-house expertise and roles (e.g. data scientist) hence embedding
personnel within water companies. Leveraging existing toolboxes (particularly
for machine learning) and workflows is becoming easier so that the barrier to
entry has now been now lowered (PhD qualifications were usually required in
the past for AI).
• Adopting open-source programming languages (such as Python) and tools
can facilitate collaboration and shared development. Software produced in
water engineering research has previously often been bespoke and stand
alone, and more thought needs to be directed at reusability, sustainability and
maintainability. For example, university research project software is usually
developed in isolation in languages such as MATLAB and may only persist
through one or two generations of researchers (such as passed down from
an academic to a PhD student).
• In the UK, centres for doctoral training like STREAM and WISE embed a
doctoral engineer in water companies helping such transfer.
• Other collaborations between industry, SMEs and research establishments in
projects (e.g. through European H2020 or UK Innovate funding) encourage
knowledge transfer. Data dives and hackathons provide forums for building
teams of collaborators.
Some private sector water utilities are already leaders in the digital space
and active in sharing their early successes with leveraging connected devices,
IoT and machine learning. This digital re-imagination of the sector will enable
a broad spectrum of outcomes from improved efficiencies to optimised asset
management across providers and potentially new business models such as
consolidated multiutility retail operations. While there is an increase of digital
adoption in water, the sector still lags behind other industries in integrating
new, smart technologies. Water utilities can benefit from the lessons learned in
other sectors (such as energy) and the established best practices and network
infrastructures. As technology evolves, the price for smart devices decreases,
the functionality increases, and consequently piggybacking on other sectors’ lead
can result in an accelerated adoption rate for realisation of benefits in the water
ecosystem.
22
S. R. Mounce