A Non-Data Scientist's Guide to Decision Intelligence

23 Mar, 2022

A Non-Data Scientist's Guide to Decision Intelligence

How engineers can impact their organizations’ decision-making processes with data-based insights.

Big data and decision intelligence

Described as “the world’s most valuable resource,” data are being generated exponentially, predicted to comprise more than 200 trillion gigabytes in the cloud by 2025.^[1,2] As a result, organizations are swamped with data at a rate faster than humans can process. At the same time, hardcore data-science skills are lacking globally, as evidenced by many businesses failing to find qualified candidates for their data needs.^[3] Access to raw data is paramount, but companies need the capabilities to derive actionable intelligence from that data.

For organizations to transform their data into actionable intelligence and create value, cutting-edge data-science technologies need to be implemented across their workforces, enabling non-data scientists to gain data-based insights. The technological research firm Gartner^[4] recognizes an emergent solution to this challenge as decision intelligence (DI), which, according to 451research.com, “promises to bring analytics to the masses. It’s all about providing insights to answer business questions, without the user requiring smarts in analytics or data.”^[5]

DI is based on advanced data-linking and automated analytic approaches, such as graph technology, automated no-code workflows, artificial intelligence (AI), machine learning (ML), and simulations, which integrate with engineering methods. With the massive increase in data and shortage of data-science skills, non-data scientists, such as engineers, can now harness this untapped opportunity in the decision-intelligence movement and improve the decision-making processes in their organizations.

Engineers at the center

Engineering investigations can yield a vast array of design solutions based on data, assumptions, and uncertainties. The growing plethora of data available to engineers, due to connected systems, Internet-of-Things (IoT) solutions, and highly instrumented assets, serves as a microcosm of what their entire organizations face. Furthermore, engineers produce massive volumes of data through modeling and simulations of ever-increasing complexity and fidelity. State-of-the-art products are requiring advanced design frameworks that may include co-simulation, integrated systems analysis, and digital twin technology, all of which demand intricate high-quality data input and generate expansive datasets. A prime example of such a system is a 3D-printed steel pedestrian bridge in Amsterdam that was recently installed as a “living laboratory,” equipped with a large sensor network. The extensive real-time data allow researchers to monitor the bridge, while a digital twin incorporates this data, fine-tuning the model over time and enabling further development.^[6]

So, engineers are at the center of this data storm. First, they are now inundated with data, such as equipment streaming live operational measurements, which are impossible to efficiently analyze with traditional methods given the accumulation rate. Second, engineers, as major data generators themselves, are part of the problem. They employ sophisticated simulations, co-simulations, and digital twins that yield an unprecedented amount of data on which decisions can be based.

Therefore, an engineer faces various categories of big data—descriptive, prescriptive, diagnostic, decisive, and predictive^[7]—which can be transformed into meaningful information with appropriate tools. Yet, despite access to this abundant commodity, as of five years ago, an estimated 70% of a company’s produced data would remain untouched, termed “dark data,” and often the used data would then be misapplied.^[8] Today, the situation is undoubtedly worse.

Decision intelligence for better outcomes

Given this technological trend, the modern-day engineer needs to incorporate data science into their methodologies, which can then inform decision-making, such that the best decisions are made as quickly and cost-effectively as possible. Doubtless, data are precious to engineers, but decisions are what move companies forward. Humans are the ultimate decision-makers, while computers perform calculations, solve equations, and execute human decisions.^[9] For optimal decision-making, we require insight on “how actions lead to outcomes,” and DI offers the possibility to combine technology with action for a better outcome.^[10]

To this end, DI introduces ways for engineers to link massive amounts of data and associated automated analytic workflows that enable them to leverage the power of data science tech without being data scientists.

Potential pathways for the engineer and decision intelligence

How can engineers take advantage of data science techniques to feed into the broader decision-making processes of their organizations? New platforms are emerging that facilitate the processing of so much data coming from different directions, easing the engineer’s workload. At the same time, they are able to transform the data into actionable information—no data science degree required. With recent advancements like graph technology and automated analytics, engineers can better leverage the data at their disposal and implement AI, ML, co-simulation, and, thus, highly accurate digital twins effectively with no additional coding.

Graph technology, in particular, which “includes graph theory, graph analytics, and graph data management,” has seen significant development recently.^[11] Contact tracing during the COVID-19 pandemic demanded an efficient way to obtain, share, and analyze a daily influx of data. The answer was graph technology, with its ability to unearth relationships among datasets quickly that might otherwise go unnoticed with standard approaches, providing actionable insight to decision-makers ranging from lawmakers to doctors to average citizens.

To take advantage of graph technology, engineers must first make sense of the vast volumes of incoming raw data, essentially impossible to do in a timely manner without automated analytic workflows, which removes the human factor from the process. Computer systems take over the data processing, including data discovery, data preparation, replication, and data warehouse maintenance,^[12] and provide high-quality datasets for the engineer to use. Using technologies like AI and ML, data trends and outliers, for example, can be discovered by this automation. An effective decision intelligence platform can eliminate up to 80% of the time wasted on data preparation, leaving far more hours available for high value–added work.

The concept of the digital twin originated with NASA decades ago during the Apollo 13 mission.^[13] Engineers now use digital twins in countless applications, including automotive, power generation, and healthcare. Slingshot Simulations defines the digital twin as a “digital environment that reflects on, mirrors, and evolves ahead of the physical environment.”^[14] In other words, this virtual model utilizes real-time data coupled with other sciences, such as simulation and ML, to provide engineers with decision-making competence—how the system operates and how it will operate in the future. Additionally, the digital twin is fueled through bi-directional data flow, receiving IoT data from sensors and supplying information based on subsequent simulations, for instance.

These analysis tools speed up the data processing and improve business viability/competitiveness. Faster and better decision-making is a result.

Further trends and challenges

DI is on the rise with an estimated one-third of major companies employing this technology by next year.^[15] Additionally, the management consulting firm McKinsey & Company predicted that advancements within the DI framework like AI may “widen the performance gaps,” with large organizations taking a considerable chunk of the business.^[16] Fortunately, data analytics is now offered as a product, providing any company with affordable access to DI tools.

DI benefits from the internal “network effects” of large data collection. For instance, Tesla is able to continuously improve its self-driving feature as it gathers data through its customer base, which increases as the autonomous capabilities are perfected. ML and simulation technologies can use the date to yield downloadable updates to drivers.^[17] Furthermore, these fast product improvement cycles keep companies ahead of their competition.

A benefit of DI is improved data intelligence. High utilization of available data avoids redundant information gathering, and a good analytics tool keeps the data available for reuse for faster decision-making. Even the digital world must strive for a sustainable future. Dark data is abundant, and the efficient use of data contributes to both economic and environmental sustainability—reduce, reuse, recycle. Surprisingly, the carbon footprint of dark-data storage is considerable, producing six million tons of CO2 yearly, while costing trillions of dollars.^[18] The speed with which the data are generated continues to increase, and companies need to acclimate with the latest data science developments in graph technology, automated analytic workflows, AI, ML, and simulation.

Data is now considered a “core system.”^[19] Organizations are moving towards data intelligence for their decision-making and investing in big-data technologies, and engineers are a significant component to the practice of DI.

Slingshot Simulations

Slingshot Simulations understands today’s challenges with decision-making, given the mounds of data coming from all directions, and equips the engineer with valuable DI tools. The company’s goal is to transform large data into valuable sources of information. Relationships within datasets are visualized. Workflows are analyzed. A limitless number of scenarios can be tested.

Slingshot Simulations has created a user-friendly application Compass: EngineTM, in which large datasets can be uploaded to the cloud for analysis. Input from various sources include real-time and historical data, ranging from public geospatial files to streamed output from an in-field product. Compass: Analytics incorporates graph technology and ML and allows the engineer to use fast automated workflows for data analytics. Non-data scientists can bypass lessons on coding and proceed to link datasets and analyze them in this low-code environment. Data processing time is reduced by 80%, leaving engineers to focus on using inbuilt and third-party simulators, as well as continuously developing their digital twin, AI, and ML projects.

DI arises through the maps and dashboards created within the Slingshot Simulations applications. Actionable insights are visualized and seamlessly shared for effective collaboration. While Slingshot Simulations is committed to sustainability through dark-data reduction, a focus on energy efficiency, and investment in environmental improvement projects, engineers are empowered with informed decision-making for improved outcomes.

About the sponsor: SlingshotSimulations

Slingshot Simulations is on a mission to enable anyone, anywhere to unleash the power of advanced data science to tackle the biggest challenges we face today – sustainability, climate resilience, decarbonisation, and more. Their technology removes the barriers to entry of cost, usability and domain expertise.

The community version of Compass: EngineTM Graph Technology Platform-as-a-Service is freely available. Its no-code environment supported by automated workflows and an intuitive user experience make it fast and easy to use, even by non-data scientists. Apps and extensions integrate our solution into your technology stack and tools of choice. What data will you make “intelligence ready” today?

References

Search for articles and topics on Wevolver

Explore topics

More by Kimberly Sweetland

Kimberly Sweetland has more than a decade of experience as an engineer in the automotive industry working on engine design, advanced vehicle technology, and turbocharger applications. She has transitioned to technical writing and editing, focusing on science and engineering. She has a degree in mech...