advertisement
advertisement

Smart machines: The 14 next big things, from instant rockets to smart tractors

IBM’s 2-nanometer superchip, robots and vehicles with a sense of touch, AI on demand, and more.

Smart machines: The 14 next big things, from instant rockets to smart tractors

Thanks to innovations in the cloud, the machines that we rely on every day are getting smarter. Increasingly, machines can “feel” the world around them thanks to a combination of sensors and cameras, then interpret that data using artificial intelligence, including computer vision. Today, our cars can detect poor road conditions and gather data that helps cities determine where to make repairs. Robots and automation make factory floors run more efficiently. The companies behind these technologies are among the honorees in Fast Company’s first Next Big Things in Tech awards.

advertisement

See a full list of all the winners across all categories here.

Winners

Azumo
For building screens that stand up to sunlight
The holy grail of mobile displays is being able to provide indoor and outdoor readability without burning through battery life. Azumo’s LCD 2.0 combines a low-power display that withstands bright sunlight with front lighting that produces accurate colors indoors. The tech could come to future phones, tablets, and laptops.

Diamond Foundry
For creating a material that can outperform silicon
One of the factors constraining the speed of current chips is the fact that silicon heats up. Diamond Foundry’s diamond material—created in plasma reactors—could be used to replace 99% of a chip’s silicon wafer, allowing it to stay cool and boosting both its performance and conductivity.

advertisement
GelSight
For helping robots develop a new sense
Researchers at MIT developed GelSight to give robots a sense of touch. The technology uses a camera that can measure distortions to its soft tactile sensor down to a micron, turning pressure to the sensor into imagery that a computer can understand. Applications include aerospace and manufacturing.

[Photo: courtesy of IBM]

IBM
For giving Moore’s law new life
Moore’s law, which says the number of transistors on a microchip doubles every two years, seemed set to end—until IBM designed a 2-nanometer chip that will let chipmakers fit up to 50 billion transistors into the space of a fingernail. Among the potential benefits: smartphones with 400% better battery life.
[Illustration: Noma Bar]

Lightmatter
For making AI faster and more sustainable
This MIT spinout’s key product is Envise, a general purpose photonic AI accelerator that brings ultrahigh performance to demanding applications such as neural networks. Since its chips run cooler, it could help data centers radically reduce their carbon footprint.

Monarch
For automating farms while preserving the environment
The compact Monarch Tractor is electric, autonomous, and can perform a wide range of farm functions (such as planting, tilling, and fertilizing). There are other autonomous tractors on the market that also aim to make farming more green, but at $58,000, Monarch’s entry in the market is remarkably affordable.

MTEK
For bringing new efficiency to factory floors
The MBrain system takes the software needed to run a modern-day factory and integrates it into a single system that can analyze each element of the manufacturing process to make it more efficient. With a digital view of their entire factory, managers also have the opportunity to reduce waste and find new ways to recycle.

Rapid Robotics
For teaching robots to do everyday work
Robotic manufacturing is typically a pricey, complex process involving custom software that costs more than the robots themselves. Rapid Robotics democratizes it with easy-to-train, on-demand robots that can handle common factory tasks ranging from injection molding to parts inspection.

Relativity Space
For 3D printing rockets in mere days
This startup devised a new way to produce orbital rockets: by printing them with the world’s largest 3D printers. Using AI and robotics, it cranks them out in days, rather than the two-plus years required by conventional means—and with 1,000 parts instead of the 100,000 required in the past.

SambaNova Systems
For letting companies train AI models in the cloud
Stanford professors created novel hardware and software for training large AI models, then made their technology easy to adopt by offering it as a service via SambaNova. The company says that its system outperforms competitors in model-training benchmarks, while using less space and power.

Skylo
For sending the Internet of Things to space
Cellular networks cover only about 20% of the world’s land mass. Skylo built an IoT platform that provides satellite connectivity to sensors located on everything from tractors to pipelines to fishing boats. In India, the technology has already helped save the lives of fishermen who might otherwise have been lost at sea.

Tactile Mobility
For allowing vehicles to feel the road beneath them
This startup’s VehicleDNA software uses sensors already embedded in vehicles to detect road hazards such as potholes and slippery driving conditions. The crowdsourced data it collects is useful to everyone from cities performing road maintenance to manufacturers designing autonomous cars.

Honorable Mentions

Groq
For designing an AI chip that can process a quadrillion operations per second

Swift Navigation
For giving vehicles satellite-based positioning that’s accurate down to 10 centimeters

advertisement
advertisement
advertisement