A More Intelligent Future
Artificial intelligence (AI) is about
to get a lot smarter and far more
ubiquitous, leading to a range of
lucrative career opportunities.
Imagine self-driving automobiles that
can converse with you, learning and
managing all the car’s functions while
passengers relax or prepare for the
workday. How about household robots
that can watch, listen, learn, and get to
know each member of the family — even choose appropriate entertainment options depending on who happens to be in the house?
This is just a taste of what’s possible with the advent of embedded artificial intelligence, the next phase in AI that promises to greatly increase performance in even the smallest devices.
AI is currently used in applications such as manufacturing robots, disease mapping, and natural language processing (NLP). But embedded AI could make a quantum leap beyond that, bringing a world of next-level AI applications into our lives, said Thomas Jannett, instructor for the Developing Embedded AI Systems certificate program.
“Embedded AI allows smart decisions to be made locally, either at the device level or at the edge of the network,” he added. “That makes it possible for the creation of new and exciting consumer, commercial, and industrial applications that intelligently process video, audio, motion, and other information at the source. Just a few years ago, it would have been impossible to run AI locally, as the hardware size and cost would have been prohibitive.” AI’s speed and capabilities have been limited by relying on cloud computing to run “computationally intensive” AI algorithms, sending large amounts of data to the cloud for processing and then back to the device.
“This back-and-forth communication, possibly over limited bandwidth links, reduces the speed at which an application can run,” Jannett said. “But embedded AI makes decisions locally, allowing a quicker response to local events so local decisions can be made much faster. It reduces or sometimes eliminates the need to send data to the cloud.”
The benefits are myriad. Along with faster response time and increased reliability and mobility, embedded AI can strengthen security and privacy. It also can lead to lower costs and allow for new, more powerful applications that are not feasible with cloud-based AI.
Robots will become much more interwoven in the fabric of everyday life, from smart automation in manufacturing to robots in last-mile delivery, potentially having a substantial impact on the supply chain. Advanced robotic systems, industrial as well as in the home, could use vision and hearing to evolve and learn more about managing their assigned tasks, Jannett said.
“Virtual assistants may make a significant transition from the voice-enabled digital assistants that we use on smartphones and smart speakers to new collaborative robots that can see, speak, understand our spoken intent and converse on a wide range of subjects, perhaps implemented as avatars. They may even talk with you while riding in your car. Imagine what can be done with an embedded AI supercomputer that is about the size of a coffee mug!”
A career on the cutting edge
The Developing Embedded AI Systems certificate program prepares students for success by providing all the tools needed to capitalize on this next major shift in AI technology. The curriculum is designed to expand career options for embedded systems professionals; software, data, and electrical engineers; data and computer scientists, and more.
Courses explore the specialized frameworks, technologies and platforms needed for creating a new generation of AI devices. Students get extensive hands-on experience investigating the advanced near real-time signal processing methods and machine learning (ML) models behind AI applications that process video, audio, and other signals.
“Students begin by studying TinyML, the field of applying ML technologies to embed AI in resource-constrained devices such as low-power microcontrollers,” Jannett said. “Then they explore the features of new and highly specialized AI hardware that might be used in consumer electronics, along with compact embedded AI supercomputers that might be used in autonomous vehicles and robots.”
Students gain a comprehensive background in deep-learning neural networks, programming in Python as well as C++ and C. In the program’s final course, students apply all the tools they have acquired to develop an actual prototype for an embedded AI application.
For professionals who want to take their knowledge and experience a step further, the Embedded Systems Engineering certificate program can be an excellent addition, providing a significant advantage for anyone who wants to master embedded AI.
“To work with embedded AI, it is important for professionals to be conversant with the ML approaches, workflows and deployment issues related to this field,” Jannett said. “Embedded systems engineering skills are critical for developing and deploying embedded AI systems.”
An autonomous future
Increasing demand is making it difficult for companies to find enough qualified candidates with the mix of essential skills needed to work in this field. More than 66,000 job openings in related fields were listed in the most recent annual EMSI survey, with more than 5% projected growth through 2029. Median annual compensation is $119,000, with highly experienced professionals earning up to $180,000 (Emsi Burning Glass – economicmodeling.com).
Expect demand to continue growing at an exponential pace, as the technology continues to evolve and expand its reach in the corporate world and at home. Embedded AI will become ubiquitous as products become smarter, more interconnected and, well, more human-like in nature.
There seems to be no limit to what this technology can eventually accomplish, becoming a part of our lives in ways we perhaps can’t envision in 2022.
“According to Jensen Huang, the CEO of NVIDIA corporation, in the future expect that everything that moves will be fully or partially autonomous,” Jannett said. “And embedded AI will be a critical part of these systems.”
Learn more about the Developing Embedded AI Systems Specialized Studies Program.