By Barry Devlin

July 2018

Fiat or Ferrari – which will your digital business need?

Autonomous vehicles are a wonderful example of how the Internet of Things and artificial intelligence are reinventing transport. They also offer a powerful metaphor of how the digital revolution is reinventing every aspect of business and IT, and driving a new architecture for information management and decision-making support.

Autonomous vehicles – self-driving cars, trucks, and buses—have caught the public imagination. From science fiction to on-the-streets pilot schemes, autonomous vehicles have become a reality in a few short years. Automobile manufacturers worldwide are competing fiercely with most planning production models with significant autonomy by 2020. The implications for drivers, car manufacturers, road planners, and society in general are wide-ranging. More on that in a moment. But there’s an even more interesting angle that applies to all industries. The emergence and impacts of autonomous vehicles offer a metaphor for digital business. From data collection to decision making, the parallels between self-driving cars and business of every shade is striking. Does your business need an architectural Fiat or Ferrari? I’ll discuss this in the second part of this article. It’s also the topic of my two back-to-back seminars with Technology Transfer in Roma, 20-22 November.

Autonomous vehicles-utopia or dystopia?
The path to vehicle autonomy began some twenty years ago when cars were required to be fitted with basic sensors and standard on-board diagnostics. Cars began to gather rudimentary data, such as engine temperature and oil pressure. With more sensors and more sophisticated on-board controllers, automobiles gained the first glimmer of sentience. They could monitor road traction to prevent skidding or wheel-locking. They could shut themselves down before being damaged by mechanical issues. Over the past few years, the information gathered and used by vehicles—location and speed, turning and braking forces, fuel consumption, external weather and traffic conditions, driving regulations, use of seat belts—has grown in variety and volume. With a combination of internally and externally sourced data, faster processors, and improving algorithms, new uses have emerged. The focus has shifted from performance monitoring and preventive maintenance of the vehicle to inferring driver behavior and supporting his/her betterment. From automated traffic management to pay-as-you-drive automobile insurance, the data gathered by today’s cars forms the foundation of new applications and novel businesses. With the transition to autonomous vehicles, the volumes and types data collected will further increase and its uses will change dramatically. Across industries and governmental agencies alike, old processes will disappear and new ones emerge. Personal driving insurance will be unnecessary. Traffic police will become redundant, driving licenses will become antiques as truck, car and bus drivers lose their jobs. Human driving for pleasure may be considered more antisocial than smoking! On the other hand, car manufacturing will become more environmentally friendly as volumes drop steeply. Car parks and on-street parking will disappear. City planning will turn from traffic management to human life enhancement. However, your complete personal travel history will become traceable by Uber-like mega-corporations and governments alike. The car that used to be the most private spot for your first hot date may become the ultimate surveillance device. This summary suggests that the impact of autonomous vehicles on society may exceed anything we can currently imagine. The same is true when we look to digital business—the widespread use of sensor data—the Internet of Things (IoT)—and artificial intelligence (AI) across all industries.

Digital business-your success or failure
From finance to manufacturing, retail to utilities, governmental bodies to charity agencies, the combination of IoT and AI (in differing proportions) will fundamentally disrupt potentially every business and IT process on which your organization depends. (For simplicity, I include government and voluntary sectors in the term “business”.) Sales and marketing based on real-time location and social media data. Market-of-one manufacturing. Distribution by drones, managed in real-time. Proactive maintenance, driven by continuous monitoring and advanced analytics. Retail without check-outs or shops with only sample stock. Middle management decisions made by algorithms. Fake news disrupting politics and democracy. The list goes on, raising vital questions for IT. Are you ready to gather and manage the data tsunami? Like an autonomous vehicle, your organization will have to handle data of very different latencies. Instant processing of huge data sets is needed to avoid collisions on the road. Reacting to potential fraud in retail or finance, or dealing with an escalating supply chain issue in real-time present similar data and processing challenges. On the other hand, you will still have to deliver legally binding, hard-to-reconcile month-end reports. In the case of automated tactical decisions, a self-driving car automatically selects the best route based on a combination of many information sources. If it gets the route wrong, you may arrive late for an appointment. However, your algorithmic decision-making systems will be taking over the tactical analyses and decisions now made by business analysts and managers. Errors in judgement may have substantial impacts—financial or legal—on the business. What governance must you have surrounding these new AI decision processes? And, as in the case of autonomous vehicles, this enormous set of inter-related data and novel processes spills beyond the boundaries of your traditional business. Online shopping expands to bricks-and-mortar. Utilities may lease housewares. Politicians become marketers. What technologies can handle such diversity?
Architecting a digital business
Whether you succeed or fail in digital business depends on your IT architecture-information, process, and people. The old ideas of separate and distinct operational and informational systems will no longer work. Choosing a single data processing technology, whether relational or Hadoop, limits your flexibility and increases risk of failure. You will need to manage data in multiple locations and systems, stationary or in-flight; access and use it through multiple tools, many of them AI-based; all the while keeping your old systems running. In short, you need both the Fiat and the Ferrari. You need a new, over-arching architecture that encompasses all data and information used by the business, all the processes-business and IT-that use information to drive actions, and the roles and drivers of the people-all the stake-holders, internal and external-for whom the business exists. This architecture is Business unIntelligence . While it has evolved since its publication in 2013 to take account of advances in IoT and AI, it remains the most comprehensive conceptual and logical architecture for the digital business. In technology terms, it bridges from the data warehouse to the data lake and creates the biz-tech ecosystem of the future.