Impact case study database

Context

Impact has been created through the commercial exploitation by energy industry multinational Schlumberger of the AI planning tools developed by Fox and Long. This has resulted in a step change in the technology applied to automatic drilling operations, with economic, efficiency, safety and environmental benefits.

Schlumberger is the world's leading provider of technology and digital solutions for reservoir characterisation, drilling, production, and processing to the energy industry. The company supplies an extensive range of products and services, from exploration through production, with a major focus on delivering hydrocarbon reservoir performance sustainably. In terms of scale, the company has product sales and services in more than 120 countries, employed approximately 82,000 people (representing over 170 nationalities) as of the end of third quarter of 2020, and reported revenues of USD32,920,000,000 in 2019.

From research to impact

Schlumberger became aware of Fox and Long’s work on planning with continuous processes through the company’s participation as an industry partner in the EPSRC Autonomous Intelligent Systems programme in 2011. Their research has subsequently been instrumental to the company’s introduction of autonomous directional drilling [ S1]. After initial evaluation, the company engaged Fox and Long as consultants from 2013, during which time they assisted with the development of an architecture for the later adoption of their methods in the application of automated drilling [ S1]. Both Fox and Long joined Schlumberger as employees in 2016.

The following impacts have been realised during the REF assessment period:

• Technology innovation, new product development, demonstration and validation;

• Improvements in well construction processes including time and cost savings, improved drilling consistency and reduced environmental footprint;

• Economic impacts including product sales and industry partnerships.

The development and implementation of AI planning supported Schlumberger’s strategic objective to move its services up through levels of automation, the scale of the reach and significance of the impact realised set out clearly by the company’s Vice President for Technology Development, Well Construction Division:

‘The real step change in our approach to automate highly dynamic and uncertain drilling process started when we engaged Derek and Maria first as consultants and eventually as full-time employees. Their work on the AI planner has become the core of our present and future automation products and services. We have established two large engineering centers, one in Houston, Texas and the other one in Beijing to engineer the research prototype that Derek and Maria developed into commercial products’ [ S2].

Impact 1: Technology innovation, new product development and demonstration

In the period 2013-2015, the POPF planner as developed by Fox and Long was trialled by Schlumberger and a demonstration system was built to explore the potential of using POPF to automate drilling in US land operations. In 2016-17, Schlumberger carried out field tests using POPF, with the intention of identifying the commercial potential of the plan-based automation of drilling.

Concurrently, Schlumberger began the construction of a commercial planning system under the direction of Fox and Long. This led to the development of a next generation planner using methods pioneered in POPF. In subsequent deployments, the new planner was used to generate the action sequences, timing, and parameters (collectively known as plans), needed to reach a given depth. The plans generated vary depending on the hole depth, trajectory, formations and other factors. Preliminary results from field deployments confirmed the potential of the plan-based automation method to speed up the drilling process and provide associated costs savings.

The underpinning technologies built on the POPF planning tool and VAL plan validation system are now embodied in the Schlumberger commercial DrillOps on-target well delivery solution [ S3] available within Schlumberger’s DELFI cognitive exploration and production (E&P) environment [ S4]. Also within the DELFI environment, the DrillPlan solution, which supports the initial well construction project planning phase, works in tandem with the subsequent automation of the drilling process via the DrillOps solution. The two solutions operate as a closely coupled pair in delivering increased automation in well construction.

Impact 2: Improvements to well construction processes: efficiency, consistency and reduced environmental risks

The implementation of AI based planning through the DrillOps and DrillPlan solutions represents a step-change in well exploration and construction. Previously, the industry standards were manual planning and a process known as scripting, in which a pre-written set of instructions, a script, inform the drilling process. Both processes come with considerable drawbacks: manual planning is time-consuming and vulnerable to human error, and scripts – although a significant improvement on purely manual planning – are usually written for a specific location, and thus lack the flexibility to be shared between drills at different sites. Both approaches can therefore lead to inconsistent drilling, delays, additional costs and errors, which may be dangerous or environmentally damaging.

Automation provides numerous benefits over both manual and scripting approaches. It results in greater efficiency and improved consistency of drilling operations. In contrast to scripting, which typically produces scripts specific to individual drill sites, automation is also a more flexible approach and thus applicable to a wider range of drill sites. This saves considerable time and allows engineers working in similar geological areas to share solutions to problems, resulting in significant efficiency improvements. In these ways the capability of automating plan generation has advanced the state of the art in automation techniques in well construction.

The transformative nature of the impact realised through the DrillOps solution is summed up by Schlumberger’s Research Programme Manager for Automation and Planning at the company’s Cambridge Research Centre:

‘The work has provided wide reaching business impact, such as through the DrillOps autonomous drilling solution, which is not incremental but marks a fundamental shift in the way we approach automation both within our own operations and in solutions we are developing with major oil and gas operators…Automation, through temporal planning enabled by the work of Fox and Long, creates consistency of operations, which drives safety, reduces time-to-target, and the minimisation of environmental footprint’ [ S1].

Specific operational performance improvements from proof-of-concept deployment in North America, the Middle East and Mexico include [ S5]:

• Improved drilling efficiency: a 20% increase in the drilling rate of penetration (ROP - the speed at which the drill bit can break the rock under it and thus deepen the wellbore);

• The ability to apply standardised procedures throughout a fleet of rigs was found to reduce non-productive time; tool failure non-productive time was reduced by 90% after the roll-out of the DrillOps solution;

• Analysis of drilling operations under manual and automatic control demonstrated that automatic control resulted in improved procedural adherence which has associated consistency and safety benefits, as well as reduced environmental footprint and risk.

Impact 3: Commercial implementation and associated economic impact

Since its commercial deployment started in 2019, the current generation of automated drilling products are used on 27 drilling land rigs [ S1], with each rig capable of drilling about 15 wells per year. The operational benefits outlined above are estimated to transfer into operating cost savings of around USD300,000 per well or USD4,500,000 per rig per annum [ S6].

In operational deployment, the step-change in operational consistency that the technology helps to deliver was reported in the Schlumberger financial results as a technical highlight of quarter 3 2020 [ S1]. This included reporting on deployment in Saudi Arabia where over 63,000ft had been drilled using the DrillOps solution, with a 17-30% improvement in rate of penetration [ S7].

In 2020, Schlumberger entered into commercial agreements with major global energy industry partners to integrate automation products including the DrillOps solution in Schlumberger systems developed for these partners. These include:

• Schlumberger and ExxonMobil, the world’s largest oil company, finalised an agreement allowing ExxonMobil to deploy the DrillOps solution as part of the deployment of digital drilling solutions around planning, execution, and continuous improvement through learning [ S8];

• Schlumberger and Chinese-based company Honghua Electric Co Ltd entered into a memorandum of understanding for the integration of the DrillOps solution into all new rigs manufactured by Honghua [ S8]. Schlumberger’s share value increased following this announcement.

Impact 4: Other applications of the planner within Schlumberger

The flexibility of the approach to AI planning, based in the Strathclyde underpinning research and embodied in the DrillOps solution, has allowed Schlumberger to implement it in other aspects of its services business [ S1].

Known as wireline operations, well measurement operations involve the deployment of sensors and associated transceivers via boreholes. These processes are typically operated manually, but this is accompanied by difficulties in controlling the winch, directing the cable and moving it a fixed speed. Schlumberger has successfully automated this process using the planner. As of 2020, Schlumberger’s system of automated wireline operations is currently under field test.

The logistics and product delivery requirements for well drilling are very large and complex, requiring a wide range of equipment and experienced staff, from a variety of locations, to all arrive at a drill site on a set day. Schlumberger has used its planner to successfully automate some aspects of this process in specific applications, with full deployment in 2020.

These new applications demonstrate the flexibility and broad applicability of the planner and its continued importance for Schlumberger’s operations.

The above body of research has directly influenced the development of the Glasgow Haskell Compiler (GHC) by providing conceptual clarifications and technical advances showing that dependent types in Haskell were both possible and practical, contradicting the previous consensus which held that an easy-to-use dependently typed Haskell was very unlikely. In 2009, McBride gave a demonstration of SHE at the Haskell Implementer’s Workshop, showing for the first time how simulated dependent types could in principle be added to Haskell [ S1]. Following this, McBride was invited to work with the GHC design team, who were keen to incorporate the features of SHE into GHC. From 2009 to 2013 there then followed regular and direct collaboration between McBride and GHC’s development team to work on GHC's type system features [ S1].

As a result of this collaboration, GHC extended standard Haskell by incorporating four programmer selectable dependent type extensions. This in turn has benefitted programming practice in Haskell by enabling greater functionality, reaching the world-wide Haskell user base. It has also led to positive economic impacts for numerous industrial users of GHC, who have been able to make use of the dependent typing features to produce significantly more reliable software with fewer bugs; as bugs take time and money to fix, this reduces the overall cost of using Haskell.

Incorporating dependent types into the GHC

Collectively, the four extensions informed by McBride’s research form the core of the dependently typed features of GHC, allowing Haskell to function like a native dependently typed language:

• DataKinds (allowing data to be used in types, released February 2012);

• PolyKinds (allowing types to abstract over the datatypes they use, released February 2012);

• ConstraintKinds (simplifying Haskell’s operator overloading by integrating it better with the type system, released February 2012);

• TypeInType (allowing more type dependency, released May 2016).

Citing R1, the paper introducing the implementation of the DataKinds extension notes that: ‘Our design is inspired by Conor McBride’s Strathclyde Haskell Enhancement (SHE) preprocessor’ [ S2a]. The coordination page for the four dependent type extensions to GHC listed above cites Lindley and McBride's Hasochism paper [ R4] as source material providing conceptual clarification on how Haskell's existing feature set can be integrated with dependent types [ S3]. Dependent types are also being incorporated into the Core language of GHC as an ongoing project, using the findings from the ‘Haskell Types with Added Value’ project as a ‘road map’, as stated by a member of the GHC development team in a 2014 keynote conference speech [ S2b]. This keynote speech also referenced R5 as a source of examples for applications of dependent types, noting that many of the examples were transferrable to GHC; the limitations of dependent types in GHC in comparison to Agda, as demonstrated in R5, were discussed as a key motivator for the integration of dependent types in GHC Core language [ S2b]. A new design for GHC with full-spectrum dependent types was published in 2017 and the specification referenced Gundry’s thesis stemming from the project with McBride in multiple places, including the statement that ‘Our design of DC is strongly based on two recent dissertations that combine type equality coercions and irrelevant quantification in dependently-typed core calculi’ [ S2c].

Confirming the role of McBride’s research in these developments, a Senior Principal Researcher at Microsoft Research, who was involved in the ‘Haskell Types with Numeric Constraints’ project and is also a member of the GHC development team, stated:

‘The effort to incorporate dependently typed features into GHC was originally inspired by discussions between myself and McBride, as well as his work on the Strathclyde Haskell Enhancement (SHE)… the project of making Haskell support richer dependent types continues to this day. These additions enable types to more precisely capture program invariants, allowing programmers to build programs that are correct-by-construction, dramatically reducing errors at compile time and hence run time’ [ S1].

Enabling greater functionality in professional programming practice

With dependent types incorporated into the design of GHC Core language, every Haskell programme compiled today using GHC uses McBride’s research. Due to the ubiquity of GHC, it is impossible to state how many programmes have relied on this research implicitly. Nonetheless, explicit use of the GHC extensions directly underpinned by McBride’s research can be identified using Hackage, which, as the main repository of open source Haskell libraries contributed by developers worldwide, provides a strong indication of how Haskell is being used [ S4]. Of the 15,433 packages on Hackage, as of December 2020:

• 998 (6.5%) use the PolyKinds extension.

• 2321 (15.0%) use the DataKinds extension.

• 1568 (10.2%) use the ConstraintKinds extension.

• 227 (1.5%) use the TypeInType extension.

• 3017 (19.5%) use at least one of them [ S4].

That 19.5% of a large repository of open source libraries use relatively new extensions to an established language like Haskell demonstrates the notability of McBride’s contribution to programming practice. Moreover, a number of these users are sizeable companies, including:

GitHub is a Microsoft-owned code hosting platform with over one hundred million code repositories. Semantic, a GitHub tool for code analysis launched in June 2019, is written in Haskell, and its codebase makes substantial use of the DataKinds and ConstraintKinds extensions [ S5]. Semantic is used to enable interactive code navigation via the GitHub web interface [ S5]. This is a deployment of significant scale, reaching millions of users worldwide.

Digital Asset, a US-based financial technology company, developed a smart contract programming language DAML in 2018 and uses the ConstraintKinds extension to simplify the code in their build systems [ S6]. A Senior Product Architect at Digital Asset has stated that ‘ConstraintKinds has been a massive simplification’ having made ‘key signatures 3x simpler, and thus more understandable’ [ S6].

Well-Typed LLP is a UK-based Haskell consultancy, founded in 2008, that is actively using and supporting the use of dependently typed Haskell in client projects [ S7]. Examples since August 2013 include the implementation of a major cryptocurrency, in which dependent types are used as part of a spectrum of verification techniques for higher assurance development, which would not be possible without dependently typed features [ S7].

Facilitating economic benefits for industrial users of Haskell

McBride's work, via GHC, has led to significant economic impacts at a number of major technology companies thanks to improved software correctness and lower development costs.

Habito is a UK-based online mortgage broker. Their online platform has brokered over GBP5,000,000,000 worth of mortgage applications since its founding in 2015, helping over 350,000 people with their home-financing needs [ S8]. Habito is almost exclusively written using GHC Haskell with dependently typed extensions, including all four extensions based on McBride’s research [ S8]. As Habito’s Chief Technologist states: ‘Heavy use of language extensions, particularly those which extend Haskell’s type system, is made throughout the Habito codebase… Over 90% of Habito’s codebase is deployed to production in customer- and institution-facing products many times a week, where it serves many thousands of visitors and customers each day. The majority of engineers interact with this codebase on a daily basis’ [ S8].

By using dependent types in their Haskell code, Habito has been able to save time and lower the risk of introducing logical errors into their programming by reducing the need to validate the same piece of data twice when checking business constraints in their refined library. Dependent types are also used by Habito to generate code that would otherwise have to be written by hand, and to automatically generate audit logs, both of which have saved considerable time for and increased performance for the business. Automatic code generation in particular has been credited as having ‘enabled changes which have increased performance by orders of magnitude, or added completely new back-ends, in relatively short timespans (weeks instead of months)’ [ S8].

Galois is a US-based research and development company. Galois has used dependently typed Haskell to develop its Crucible software, a retargetable software simulator for analysing software written in multiple programming languages, primarily for security and correctness. Crucible makes essential use of dependently typed Haskell to enforce crucial invariants about the systems it is analysing, preventing bugs in the analysis [ S9]. Galois has calculated that Crucible has played an important role in contracts worth over USD22,600,000 (10-2019), demonstrating its significance to the company [ S9]. These include use by Amazon to help ensure security of their ‘s2n’ cryptography library and in several projects funded by the US government on Fully Homomorphic Encryption and Software Brittleness.

As of 2019, US-based technology company Google uses dependently typed Haskell to develop a project to produce a hardware artefact and its supporting software infrastructure that will, in the words of two Senior Software Engineers at Google, be ‘deployed at a huge scale within Google and serve high volume traffic’ [ S10]. This project is being developed with a team of 20 engineers, ‘a large team of engineers (even by Google standards)’ [ S10]. Language extensions that support dependently typed programming in Haskell are used extensively in this project: by January 2019, the codebase contained 15 modules that use the TypeInType extension, 378 modules that use DataKinds, 51 that use PolyKinds and 69 modules that use ConstraintKinds. Highlighting the benefits of these extensions for the project, the Engineers stated:

‘Since we are very much a product group required to develop and deploy sophisticated software in a tight timescale we rely heavily on Haskell’s expressive power as well as the robustness of language extensions for features like dependently typed programming, many of which have been developed at your department [Computer & Information Science, University of Strathclyde]…These language extensions have significantly increased our productivity and helped to find errors early at compile time…Given the tight deadlines we operate to, the extra productivity afforded by the dependently typed programming features developed by Dr McBride et al. has been of great operational importance to us’ [ S10].

The Test of Change approach used by DHaWG in the three programmes discussed (dallas, ScotCap and the National Atrial Fibrillation Programme) ensured that implementation and evaluation were embedded throughout the programmes, from procurement through to implementation, and that evidence was generated to support the adoption of new digital health services within the local context. Real life and real time experiences with patients and clinicians were also captured in order to provide usable insights for others to inform their adoption of services within different clinical sites and contexts. By doing so, Lennon, McCann and Macguire’s research has:

• Informed healthcare policy by raising awareness of digital health and real-world evaluation in policy debate, and through direct advice to the Scottish Government;

• Influenced the delivery of healthcare practice and services, leading to an improved patient experience;

• Informed expansion of the digital health industry, by increasing visibility and supplying evidence for deployment business cases.

Informing healthcare policy

The dallas programme reached over 169,000 consumers, businesses, patients and healthcare professionals during its three-year lifespan, and has been repeatedly cited as having significantly raised national awareness of both digital health and real-world evaluation methods throughout the UK [ S1- S3, S6, S7]. Director of Innovation and Deputy Chief Executive of the Digital Health and Care Institute (DHI) confirms that dallas ‘engaged and inspired others by providing a large scale working example of “how to do innovation in a complex environment”’ [ S2]. As a result, the findings of the project, in particular the recommendations stemming from it [e.g. R3], Lennon was invited to present the key findings to the Scottish Government’s Digital Health and Care Strategic Oversight Group and lead a workshop on how dallas might inform the key strategic priorities for the Scottish Government’s Digital Health and Care Strategy [ S3]. Head of eHealth Strategy and Policy for the Scottish Government highlighted how ‘these recommendations resonate with the Scottish Government’s direction in this area and the workshop and presentation given by Dr Lennon were extremely timely in helping to ensure that Scotland continue to be among the global leaders in the future of digital health and care’ [ S3]. R3 was also cited by the Scottish Government in 2018 in two national policy documents on the future strategy for digital health and care delivery in Scotland; for example, the research was used in the Technology Enabled Care: Data Review and Evaluation Options Study (2018) to highlight issues related to implementation of digital health platforms, including ‘disempowerment felt by health staff, inequalities of access because of different levels of digital/IT skills, and fears around data safety, especially where platforms are run by non-NHS bodies’ [ S4]. Continuing this engagement, Lennon was invited to present these recommendations to the Scottish Government’s Digital Health and Care Strategy Advisory Board in 2018, and these recommendations have since been used to directly inform Scotland’s Digital Health and Care Strategy [ S5].

Influencing the delivery of healthcare practice and services

The evaluation methods developed in the underpinning research have provided the NHS Scotland and other healthcare providers with quicker and easier ways to evidence and procure new digital health technologies. Indeed, the work of DHaWG in documenting and evaluating examples of digital health innovation in complex environments has been acknowledged as contributing significantly to increased digital innovation in Scottish healthcare over the past five years [ S2, S6]. As the Director of Innovation and Deputy Chief Executive of DHI confirms, the dallas programme ‘influenced the integration of design-led innovation and citizen engagement as fundamental components of health and care service redesign in Scotland’ [ S2]. Components of this have also been adopted by regions in Spain, the Netherlands and Norway as part of their service transformation agenda [ S1].

DHaWG’s report on the National Atrial Fibrillation (AF) Programme [ R5] was used by NHS Lanarkshire and the Scottish Government to assess how AF screening devices can be scaled nationally to improve AF screening and reduce the number of strokes. As highlighted by the CEO of DHI, funder on the Programme:

‘[Lennon and McCann]’s engagement with NHS Lanarkshire in support of the service redesign in the detection of atrial fibrillation was a key factor in business case development to improve national screening service using digital technologies. This will improve the patient experience, reduce demands on in-patient beds and, in the medium to long term, improve outcomes for patients by reducing the risk of profound disability or death through stroke’ [ S1].

Likewise, the DHaWG evaluation conducted for the ScotCap programme [ R6] formed the foundation for a business case on colon capsule endoscopy (CCE) and enabled the industry partners to improve patient experience by responding to feedback in the evaluation to enhance the comfort and wearability of the recording device and belt worn to transmit images from the capsule [ S7]. CorporateHealth International, a Danish Company and partner in the ScotCap project whose specialists provide readings of images from the video capsule, confirms that: ‘The service evaluation phase of ScotCap created the foundation for the roll-out of CCE in Scotland’ [ S8]. By December 2020, ScotCap had been launched in NHS Highlands & Islands, Greater Glasgow and Clyde, and Tayside (comprising 35% of the Scottish population) [ S1, S7, S8]. By reducing the need for in-hospital colonoscopies across Scotland, use of ScotCap significantly reduces waiting times for bowel cancer screenings [ S7]. Speaking in media coverage at the time, Chief of Medicine for NHS Greater Glasgow and Clyde stated: ‘This exciting development will help cut waiting times and will mean that many patients will avoid the need for more invasive tests’ [ S9]. Similarly, the NHS Lanarkshire Associate Medical Director summarised the benefits of the new service: ‘It enables us to conveniently explore the entire colon and can help us to detect or exclude cancer more quickly, as well as reducing the waiting time for colonoscopies’ [ S9]. This impact has even more significance since the COVID-19 pandemic in Scotland, which has led to widespread delays to health and care practices and the need for telemedicine has become more apparent, as the Managing Director for CorporateHealth International highlights:

‘Even though the pandemic complicated the administrative side of the roll-out, the shortfalls of traditional methods have been broadly exposed. To efficiently diagnose patient for GI symptoms, we expect that 5000-10,000 patients will be using the managed service in the next 9 months. Cancers that would have stayed hidden until the waiting list had been cleared will have been detected’ [ S8].

A collaborating partner from NHS Highlands confirms that, as a result of DHaWG’s evaluative work, ‘new service models are being adopted… and patients have a greater choice of technology for improving their quality of care’ [ S7].

Informing expansion of the digital health industry

For small to medium-sized digital health enterprises, participation in these three digital health implementation programmes has increased their visibility among potential healthcare clients and other clients, and established connections which have led to commercial growth for industry partners [ S6]. As the CEO of DHI makes clear:

‘DHaWG are working with DHI and procurement teams in the Scottish Government, NHS National Services Scotland and the enterprise umbrella organisation Connected Scotland to allow Scotland to ‘procure innovation’. Strathclyde are helping pioneer this ‘Innovation Partnership’ approach, allowing a competitive R&D process to result in the setting up of a procurement framework, meaning those businesses contributing to innovation have a clearer route to large-scale deployment of digital health technologies’ [ S1].

This acknowledgment is echoed by CorporateHealth International, who described their Innovation Partnership as ‘pivotal’: ‘the group from DHI and Strathclyde provided scientifically sound patient and service insights. This together provided the evidence needed to convince stakeholders from all functions in healthcare, government, academia, and industry to move forward with the roll-out’ [ S8]. As a result, businesses are now realising faster routes into the care market [ S7].

As an example of this, Sitekit Ltd, a Scottish small and medium-sized enterprise (SME), collaborated with Lennon on the dallas programme, during which time Lennon’s research evaluated barriers and facilitators to the adoption of digital personal health records, which was one of Sitekit’s core interests [ S10]. This provided Sitekit with the necessary know-how to identify commercial opportunities in the emerging Digital Identity market and ultimately set up a new business division within Sitekit for this purpose. Sitekit has since launched the eRedbook, a digital record of a child’s growth charts, vaccination data and other NHS health records, which was launched in Liverpool in 2014 and in London in 2017. By late 2020, an estimated 30,000 children had held an eRedbook account [ S10]. In 2019, Sitekit became a development partner in the development of the ‘Share2Care’ Local Health and Care Records (LHCR), a collaborative programme in North West England NHS and Local Authorities, which sees health and care records shared electronically between health and care providers. Currently there are 11 NHS organisations and 9 Local Authorities using the Share2Care LHCR [ S10]. In 2020, Sitekit became the prime contractor for the NHS Covid-19 Digital Staff Passport, which enables NHS staff to move more safely and efficiently between NHS organisations to support the COVID-19 response and is used by 67 organisations in NHS England [ S10]. Through the development and delivery of digital identity services, Sitekit has grown from 21 employees in 2014 to 72 employees in 2019 and has increased in turnover from GBP1,564,636 in 2014 to GBP8,413,604 in 2019 [ S10]. It has delivered digital identity services for NHS, National, Regional and Local Government and the private sector, and in 2019/20, Sitekit’s digital identity division was contracted by a major multi-national client in the finance sector to build the prototype of a global-scale digital identity service, which is currently in pilot in Australia [ S10].