The "edgification" of ML into sensors and microcontrollers is just beginning - there is no end in sight to defining creative use cases.
Millions to Billions – Millions of TinyML (microcontrollers) capable devices are shipping now, and shipments will grow to 2.5 billion annually by 2030.
Embedded devices are used widely across a diverse range of fields including consumer electronics, wearables, public safety, industrial, and wildlife conservation, among others.
USE CASES OF Embedded ML within the enterprise
Tracking real-time health metrics; Biomonitoring of soldiers and first responders – predicting their core temperatures and heat stress.
Circular economy models for Supply Chain; Brambles UVP is tracking the location and use of 330 million pallets worldwide in 60 countries, renting them out to customers and collecting them back. Working on future uses cases of serialization and motion detectors.
WITH GREAT POWER COMES GREAT RESPONSIBILITY
How to build scalable deployment techniques for Embedded ML when the number of devices in production grows from tens to tens of thousands.
More trustworthy measures for inference accuracy must be developed via ML software engineering.
MLOps must adapt and evolve its operations to suit machine learning models.
Last but not least - Earning the trust of enterprise customers will require education and training around the testing and taking to market of data-driven engineering products and services.
Momenta is the leading Digital Industry venture capital firm accelerating digital innovators across energy, manufacturing, smart spaces, and supply chain. Led by deep industry operators across its venture capital, strategic advisory, and executive search practices, Momenta has made over 50 investments, with notable exits to SAP, PTC, and Husqvarna Group. Schedule a free consultation to learn more about our Digital Industry practice.