What technologies for the data-driven era are emerging now? As you might guess, much of the activity seen across the market is connected to AI. At the same time, there is a critical data element that needs both to support AI implementations, as well as be supported by AI.
DBTA queried industry leaders to get their views on the most compelling technologies now emerging in this data-driven era.
AI-POWERED DIGITAL TWINS
Emerging digital twin technology—which enables the digital replication of systems, processes, and facilities—brings together AI, the Internet of Things (IoT), and operational data to help users analyze, simulate, and optimize complex processes in real time. In the process, digital twins are evolving “beyond static models into dynamic, AI-driven environments where manufacturers can test strategies in simulated worlds before applying them in the real one,” said Glynn Newby, strategic advisor for manufacturing at SAS.
Benefits: AI-powered digital twins provide executives with “a deeper understanding of how their operations perform and where inefficiencies exist,” Newby continued. This has positive connotations for process optimization, product quality, and equipment utilization, as well as reduced energy consumption.
Digital twins can play a role in “routing and scheduling automated guided vehicles,” for example. “This technology helps companies shift from manual oversight to intelligent orchestration.”
Adoption issues: Data integration is the most pronounced challenge to developing digital twin technology. “Digital twins depend on bringing together data from many different sources across the enterprise, including machines, sensors, environmental inputs, and production systems,” said Newby. “If that data is siloed, incomplete, or inconsistent, the value of the model is limited.” Along with data integration, companies need to prepare their people and processes for AI-driven decision making.
AI AGENTS FOR INTERNAL DATA
AI is evolving to be a serious partner in human decision making, and this requires agentic AI that can reason over an organization’s own trusted data, not just the open web, said Elizabeth Ngonzi, board member with the American Society for Artificial Intelligence.
Benefits: “These agents can turn scattered, siloed information into timely, explainable insights,” said Ngonzi. “They can scan large volumes of reports, policies, transaction data, and communications, then surface patterns, risks, and opportunities that humans might miss or would need weeks to uncover.”
This helps humans avoid spending their days hunting for and stitching together information.
Adoption issues: Any agentic AI effort pertaining to internal data must start with very human questions, said Ngonzi. “What pain points are we solving? What are our goals? What constraints and capabilities do our people actually have? If your data is messy, biased, or poorly governed, agentic AI will simply amplify those problems at scale. And if teams are not trained and empowered to question the agent’s output, you risk quietly automating bad decisions.”
EDGE ANALYTICS IN IOT DEVICES
Edge analytics in IoT devices, especially surveillance cameras, has come a long way in the last 12 months. Relatively small devices can now store data to supplement video images, said Fredrik Nilsson, VP of the Americas at Axis Communications. “For example, within just one surveillance camera, a model can run that sends not only the license plate, but also car make, model, and color. This means that security cameras no longer need large operating systems to recognize most vehicles, humans, colors of clothes, and safety gear, while simultaneously sending more data points than ever before.”
Benefits: Video devices are typically associated with security, but surveillance tech can also enhance business intelligence, said Nilsson. “For perspective, today, there are an estimated 100 million security cameras. Most videos recorded are stored for 20 days, then deleted, and it is very cumbersome to analyze every recording for business and security insights. Today’s AI enables cameras to extract relevant data for business intelligence and operational efficiency, which can be easily fed into internal or external LLMs to manage business operations.”
Adoption issues: Educating the market is a challenge, as many organizations don’t always understand the specific capabilities video technology offers, said Nilsson. “Security departments tend to be responsible for implementation—even though they may not understand exactly how other departments are hoping to leverage them.” Greater education and collaboration across organizations are needed, he added.