Five recommendations to enable next-generation R&D in the cloud

A blog post by Diana Kearns-Manolatos, senior manager, Center for Integrated Research at Deloitte.

The pandemic has forced organizations to innovate products and services and to rethink existing R&D strategies. Organizations have had to look for new ways to share information with privacy in mind and collaborate globally—at unprecedented speed. While this change is happening across almost every industry, the stakes have been particularly high in the life sciences and health care industries. Organizations have needed to quickly advance public, private, and academic partnerships in order to manage everything from contact tracing to the search for therapeutics and a vaccine to support COVID-19. Over the course of this year, the global community has made tremendous progress in facilitating data-sharing and coordinating across organizations in new ways previously thought impossible, with cloud technologies as a key enabler.

The cloud brings unique advantages that are essential to a next-generation R&D strategy.

• First, cloud-enabled data platforms allow organizations to share information through a number of different approaches: enterprise data management; external data exchanges; open and interoperable API-driven architectures; and cloud data lakes, data warehouses, and data storage, to name a few.
• Second, cloud provides the infrastructure needed to scale up and down usages as needs change, as well as a unique ability to create a global, networked ecosystem for knowledge-sharing. For example, a scalable cloud ecosystem can bring together academic research, clinical data, real-world hospital data, and private-sector resources into a single network to accelerate the entire ecosystem’s ability to improve early diagnosis of diseases and support real-time, personalized health care.
• Third, cloud services allow organizations to tap into value-added resources to advance their R&D and innovation work, including analytics, IoT, and AI/ML services (cloud machine learning). For example, one large oil and gas company was able to use a range of cloud services to collect data from the well head, feed that data into advanced commodities trading models, and provide better real-time decision-making across an increasingly complex energy supply chain.¹

And for each of these advantages, there is tremendous optionality and flexibility with how organizations choose to share, secure, and act on their data—whether to securely share clinical data and real-world evidence from health care patients or to support pharmaceutical research, bioinformatics, or advanced health care analytics. While Deloitte’s research has shown 60% of US life sciences and health care (LSHC) organizations host more than half of their applications on the cloud already,² the next frontier will be about facilitating greater data interoperability within organizations and across organizational boundaries.

To follow are five key recommendations LSHC organizations, R&D professionals across industries, and CTOs can keep in mind when establishing a next-generation R&D strategy powered by the cloud.

1. Build a business case for data-sharing. While data-sharing is essential to build on the most robust knowledge supply available, organizations need to understand their incentives for sharing data and feel confident that the correct data security controls are in place to protect data ownership, intellectual property, and commercial incentives. Part of the reason LSHC organizations have been able to quickly make progress with cloud-enabled data infrastructures is that the shared global mission to rally talent and resources around a search for therapies and a cure for COVID-19 created a clear business case for data-sharing. Additionally, organizations will need to think about whether a migration, new infrastructure, or cloud-native development is the best approach based on the unique business case. Cloud-native application development, for example, may be the best course of action to create a digital application for contact tracing, COVID-19 testing, or a number of other scenarios that require massive data management across a large number of users practically overnight.
2. Choose a fit-for-purpose data-sharing model. Next, it’s important to understand that data standards may vary considerably when collaborating across teams and organizations, and therefore, data interoperability can be a challenge. At the outset, a F.A.I.R. data approach will create a strong foundation to make sure that data is findable, accessible, indexable, and reusable (F.A.I.R.). With that foundation, organizations should then understand that a shared, centralized data warehouse that requires standardized and interoperable data is not the only infrastructure option. Cloud-enabled data infrastructures can come in many forms, including enterprise data management, data exchanges, API architecture strategies, the centralized data warehouse, and data lakes. For a single organization, this may enable them to take data saved on local hard drives (such as laboratory data) to make that data searchable as a reusable asset across teams within the organization. In terms of data-sharing, this may include tapping into data exchanges like those bringing together clinical data and real-world evidence that cloud providers offer to get access to new commoditized data. For highly regulated industries such as pharmaceutical companies, this may include sharing noncompetitive, HIPAA-compliant placebo data. In other cases, this may be a matter of thinking of greater standardization across application programming interfaces (API) to allow for more standardized infrastructure should the organization choose to share information at some point in the future. There are a range of options, and while cloud data warehouses do provide unique scalability benefits and the ability to ingest petabytes of data, they’re not the only option.
3. Address data privacy. One of the barriers that has prevented organizations from embracing the cloud has been privacy concerns; however, when configured properly and with necessary access controls and identity management in place, the cloud is highly secure. For data privacy, look at federated security across the cloud network; establish a clear controls framework that obfuscates, anonymizes, de-identifies, or pseudonymizes data; and implement identity management access controls. For more, see our recent Federated Security post.
4. Take advantage of the full network and scalability potential of the cloud ecosystem by implementing the right cloud engagement model for your organization. Some organizations have opted for more service-oriented operating models powered by SaaS. In these situations, organizations use software to help with better coordination and collaboration powered by the cloud. During the pandemic, these types of solutions have received greater attention as remote work has prompted colleagues typically working in the same office to use virtual collaboration tools. Those tools can help support collaboration beyond the day-to-day team by providing a solution for knowledge-sharing, IP tracking, financial reporting, and more. From an infrastructure perspective (IaaS), other organizations have created innovation development sandboxes to spin up and down infrastructure for fail-fast scenarios. Still other organizations might choose a data-oriented approach delivered across shared platforms (PaaS) or even process-oriented approaches. Whatever the approach, think beyond your own organization, for this is where the network effect of cloud technology comes into full force.
   
5. Understand cloud services that can deliver added value. Cloud providers have a growing set of services across analytics, AI/ML, and IoT that can help organizations address their innovation time, talent, and technology needs. LSHC organizations are using these services, for example, to accelerate clinical trial recruiting, perform advanced diagnostics for early detection of cancer and rare diseases, and more.

As organizations look to pivot their business strategies and develop technology programs that support net-new innovation or to advance information-sharing and collaboration across existing R&D programs, cloud can be a key enabler. Taking advantage of that opportunity will require shared understanding of the different operating models available and aligning the best one with the organization’s (or group of organizations’) individual and shared goals. Done right, businesses and governments can accelerate the R&D and innovation initiatives in a way that balances commercial incentives and supports the public good. For those looking to take that first step, examine your current-state technology environment, and think about your ideal future state as a first step.

For more actionable recommendations, and to learn how specific organizations are adapting to adjust their R&D strategies in the cloud, read our article: Innovating R&D with the cloud: Business transformation could require cloud-enabled data, ecosystems, and services.