From Speak & Spell to Edge AI: How TI’s Early DSP Innovation Shaped Modern Intelligent Devices

Edge artificial intelligence (AI) holds tremendous promise for transforming everyday life. In wearable cardiac monitors, for example, edge AI allows devices to process heart rhythm data locally, remove noise, and detect irregularities in real time—reducing healthcare costs while expanding access to care. In HVAC systems, edge AI can learn occupancy habits and automatically adjust temperature settings according to weather conditions, improving both comfort and energy efficiency.

Remarkably, the roots of these sophisticated capabilities can be traced back to a children’s learning toy.
In 1978, Texas Instruments introduced Speak & Spell, an educational device created to teach spelling and pronunciation. The toy featured a built-in speech synthesizer powered by the first chip in the industry to incorporate digital signal processing (DSP) logic. This breakthrough in DSP made it possible to process real-world signals efficiently and in real time directly on a device—principles that now underpin many of TI’s modern edge AI solutions.

Roland Sperlich, Vice President and General Manager of Processors at Texas Instruments, later sat down with Gene Frantz, one of the pioneering engineers behind the DSP technology used in the Speak & Spell speech software. Before retiring, Gene also served as a Senior Fellow in TI’s Technical Ladder program. In their discussion, they reflected on:

• How the Speak & Spell team used creative approaches to secure resources and funding for development

• The technical challenges of implementing real-time speech processing with linear predictive coding, selecting appropriate voice output, and determining the device’s vocabulary

• The logistical difficulties of coordinating integrated circuit design, speech research, and product development teams located across different sites in Texas

• How the team worked closely with TI leadership and end users to refine and improve the product

• The evolution of TI’s DSP technology into today’s edge AI applications across automotive, medical, and industrial automation sectors

• Key lessons learned from early DSP development, including the value of giving engineers autonomy and ownership over their projects