Robots for hire: Three hardware technology growth trends

As a hardware and frontier technology banker and enthusiast, I am fortunate enough to get a front-row seat to the world’s most incredible inventions, from low-orbit satellites and sailing drones, to robotic beehives. But despite the wow factor, hard tech and deep tech are still at the fringes of the innovation economy. Too many investors consider physical machines hard to build and slow to scale, but that’s beginning to change.

An evolution is underway in which software-enabled machines proliferate like never before. The reshoring of US manufacturing, advancements in computing, and the accessibility of low-cost hardware and infrastructure are tipping the scales in favor of widespread adoption of technologies that were once science fiction. Propelling this trend is a subscription-based sales model referred to as Hardware-as-a-Service (HaaS), borrowed from software-as-a-service, enabling hardware automation in new sectors — in manufacturing and beyond.

We recently partnered with Eclipse Ventures to publish our first-ever State of Hardware-as-a-Service report. The report looks at the HaaS model to identify the key indicators every founder should know. If you haven’t had a chance, watch the supporting webinar, The Metrics that Matter in the Evolution of Hardware –– it features a panel of full-stack industry experts and founders discussing the KPIs and themes that are driving industry leading hardware companies forward.

The report does a great job providing a deep dive into the mechanics of the HaaS model, but only scratches the surface on the macro themes driving its emergence. This blog post expands these forces and explores some of the “ah-hah” moments we discovered when creating the report.

1. Reshoring is driving demand for automation

US companies have been moving factories back home for over a decade, but that trend is ramping up. Global uncertainty caused by the COVID-19 pandemic and worsened by the war in Ukraine has pressured executives to de-risk their supply lines. CEOs on earnings calls are mentioning “onshoring,” “reshoring” and “nearshoring” more than they did during the height of the pandemic (and 10x more than before its onset).


CEOs on earnings calls are mentioning onshoring, reshoring and nearshoring more than they did during the height of the pandemic (and 10x more than before its onset).


Companies are making good on these promises. Last year the US added 10,008 factories, a record-high and more than double the 4,341 added in 2020, according to the Bureau of Labor Statistics. The growth has continued in 2022. Factory groundbreakings were up 170% through July, and manufacturers spent $92 billion on new plants in the 12 months ending in June, up 20% from a year ago. Government incentives are funding massive projects across the country –– three semiconductor plants in Arizona, an electric vehicle “mega campus” in Tennessee and a steel mill in Arkansas. These plants will require machines to operate them. To compete with low-cost labor countries in Asia, US manufacturers must rely on automation to manage expenses.

US manufacturing bounces back

Notes: Annual establishment count in Q4
Sources: BLS and SVB analysis.

Technology enables fewer workers to produce more goods. Take agriculture, for example. Farming became so mechanized in the 20th century that the share of agriculture workers fell from about half of the US labor force in the late 19th century to just 2% by 2000, based on analysis of US Department of Agriculture and US Census Bureau data. The advent of combine harvesting allows for mass harvestings of crops such as corn and wheat; however, not all crops are suited for combine harvesting. Raspberries, asparagus, cherries and most of the fresh produce on grocery store shelves is still picked by human hands, too delicate for heavy machinery. Robots are targeting these labor-intensive markets. For instance, an automated grape harvester was introduced in 2001 that massively disrupted the raisin industry. The percentage of grapes picked by machines jumped from 1% in 2000 to 45% by 2007, according to a USDA report. Today, agriculture startups equip machines with AI-visioning software to pluck the best strawberries and other hard-to-pick produce.

The same trends that have changed farm work are also changing factory work. Forty years ago, when US manufacturing was first being outsourced, automation was only viable for the most expensive processes. The earliest commercial robots used were in heavy, durable goods such as automobile and airplane manufacturing. Now that low-cost automation is possible, hardware tech companies are extending factory automation in ways that weren’t feasible until now.

Formic Technologies, a venture capital-backed company founded in 2020, offers industrial robots for tasks like welding, machine tending, inspection and handling. Instead of a one-time sales model, Formic uses a HaaS subscription model and charges companies by the hour of usage, the same way hourly workers are paid. This approach appeals to small- and medium-sized manufacturers who are deterred by the upfront capex costs of purchasing the machines. This HAAS model is not entirely new, but rising wages and lower costs of technology are making the costs of a robot more competitive with human labor. For instance, the average pay for a US manufacturing worker is $30.82 per hour. Formic rents its robots for as little as $8 per hour.

2. Advances in technology are driving new uses for robots everywhere

While reshoring creates more demand for industrial automation, this shift is not confined to a single industry. Nearly every industry is adapting to automation. A decade ago, hardware founders were typically roboticists building complex machines from the ground up. Today, however, hardware components have become so accessible that software engineers can develop applications on simple devices to serve a niche customer base or access a particular data set.

SafelyYou is a good example of hardware providing a solution for a problem that software alone couldn’t solve. CEO George Netscher, who participated in the SVB-sponsored HaaS webinar, founded the company to apply AI-enabled video technology to address one of the most common safety issues for people with dementia — falls in residential buildings. Netscher built the prototypes using webcams and Raspberry Pi computers. Five years later, the company has already detected 35,000 falls and prevented thousands more.

“The only way we could serve these folks in a meaningful way was through hardware,” SafelyYou CEO George Netscher said.

For HaaS, the ability to add features to a customer’s existing devices with a simple software update unlocks higher earning potential over time. For example, Cobalt Robotics took this approach to add functionality to its existing products — security robots that patrol office buildings. At the start of the pandemic, Cobalt’s clients needed new safety protocols for employees and visitors to enter their buildings. So, the company added features to existing hardware to take temperatures, enforce mask policies and monitor social distancing.

Like Cobalt, most of today’s hardware startups aren’t developing cutting-edge technologies. Instead, they’re applying existing tech (like machine vision) to new use cases (like office security). CEO and co-founder Travis Deyle calls this approach “operational money ball,” allowing companies to “take what is already right there ready for deployment, package it up and be able to deploy these services,” he shared during the HaaS webinar.
Startup founders no longer must out-innovate the multimillion-dollar machine in the center of the factory — they now enhance or expand it, leading to bottom-up sales strategies and opportunities to target operating expenses rather than capital expenses.

An excellent example of this is from Third Wave Automation. The company automated the forklift by adapting to existing factory floor plans so companies don’t have to make expensive changes. Before founding the company in 2018, CEO and co-founder Arshan Poursohi tested the idea by pitching it to factories door-to-door in California. “When we said we were going to automate the forklift, everyone said, ‘Great, how many can I get?’” said Poursohi, another webinar panelist. Third Wave raised a $40 million Series B round last year, one of 355 VC deals totaling $8.8 billion for robotics companies, a record high.

After years of skepticism toward hardware and frontier tech companies, investors are pouring more money into physical technology. Part of the reason for this is the rise of the HaaS business model.

US VC investment in robotics companies & First VC round deal size by sales model

Sources: PitchBook and SVB analysis.

3. A new payment model is making hardware tech more investible

Over the last few decades, VC investors have generally shied away from hardware technology. “Hardware is hard” has become a well-worn cliché among founders and investors alike. It’s true that compared to software-only companies, hardware requires more startup capital and offers lower initial margins.

However, there are advantages to building physical things. Hardware customers tend to be stickier. Just imagine having to rip out the equipment in your factory to change providers. That’s more difficult than switching to a new HR software. Physical infrastructure is an excellent moat against the competition. Apple, Amazon, TSMC, SpaceX and Tesla all remain ahead of their competitors because of the physical footprint necessary to deploy their solutions at scale.

The HaaS model combines the scalability of software sales with the stickiness of hardware. Given the current market conditions, many companies are cutting capex spending. Intel, for example, is cutting capex by $4 billion this year. HaaS companies stand to benefit from this shift by transforming machinery from a heavy upfront capital expense to a cost-cutting operational expense. It can also simplify the sales process. While capex generally requires upper management approval, a monthly operating expense is generally up to a mid-level manager. This lower barrier creates an easier sales ramp and may lower acquisition costs.

Key metrics by business model

Cash flow by business model: HaaS vs. traditional hardware

The robots we already know

The “Made in America” premium has faded away over the last 30 years. Companies now see having control over their supply chains as an insurance-like cost. They’re willing to pay more for that control, but not much more. Therefore, automation will have to bridge the gap by making production costs competitive with low-wage foreign markets. Hardware startups are rising to meet that challenge just as technological advancements lower founders' barriers. These forces are resulting in new use cases for hardware tech companies.

You don’t think of your washing machine as a robot, but maybe you should. Cobalt’s Deyle says robots are what we call machines that haven’t reached mass adoption yet. Once they’re incorporated into our lives — automating tasks like washing our dishes, harvesting our wheat or vacuuming our floors — we call them by the thing that they do rather than by the name “robot.” With macro forces pushing automated machines into more uses, hardware and frontier tech companies are poised to occupy a greater role in our lives.

Where does SVB come in? Supporting innovative companies is the crux of what we do — helping founders and CEOs with the financial solutions, expert guidance and a network to bring their innovations to life at all stages of growth. Learn more.