7 startup lessons from Boston robotics veterans

WALTHAM, Mass. — While there is plenty of interest in robots, drones, artificial intelligence, and autonomous vehicles, the path to startup success isn’t always easy or clear. Robotics experts shared some of their hard-won experiences at an “Ask Me Anything” event that Vecna Robotics hosted at its new headquarters here last week.

The Robot Report Editor Steve Crowe moderated the panel, which included the following leaders in the vibrant Boston robotics scene:

• Christopher Cacioppo, co-founder and chief technology officer at 6 River Systems Inc., which makes collaborative mobile robots and systems for warehouse automation
• Leif Jentoft, co-founder, chief strategy officer, and head of partnerships at RightHand Robotics Inc., which makes piece-picking systems for e-commerce order fulfillment
• Tom Ryden, executive director of MassRobotics, an organization dedicated to serving the robotics ecosystem in Massachusetts
• Daniel Theobald, CEO of Vecna Robotics, which makes autonomous mobile systems for industrial use, and co-founder of MassRobotics
• Clara Vu, co-founder and vice president of engineering at Veo Robotics Inc., which is using sensors and AI to make industrial robotic workcells safer

Here is some advice they had for robotics startups:

1. Consider Massachusetts for a robotics startup

The Boston area has a “critical mass” of talent, agreed the panelists.

“The rich ecosystem includes lots of robotics companies, colleges, and universities,” said Ryden. “There are 36 robotics laboratories here, as well as lawyers who can help a startup with intellectual property issues. Venture capital firms are also familiar with the area. The community is important for finding a mentor.”

“People may not know that New England has a lot of robotics suppliers,” noted Vu. “There are design and mechanical shops and contract manufacturers.”

In addition, while Silicon Valley is known for its software expertise and funding availability, Massachusetts has a tradition of interest in practical solutions dating back to the first Industrial Revolution, said Theobald.

2. Find the right problem to solve

“Novelty doesn’t sell anymore. It’s not enough for robots to be cool,” said Vecna’s Theobald, reiterating a point he made in a recent column in The Robot Report. “They have to provide value and an immediate return on investment. Why start with a problem that others have solved or the hardest one, like self-driving cars?”

“No one thought we’d someday use a computer to order a coffee,” he said, referring to smartphones. “You have to choose the right problems, and there are lots of opportunities for automation.”

RightHand’s Jentoft agreed, saying, “What can we sell that will solve what customers need solving? For us, that was reliable fulfillment.”

“When I looked around, there were a lot of tech startups without a business plan,” said 6 River’s Cacioppo. “You need to know the industry you’re serving. We didn’t want to do grasping, which others like RightHand are already doing well. Try to pick only one rabbit out of your hat.”

“Co-founder Patrick [Sobalvarro] got the idea for Veo while visiting a factory,” Vu recalled. “A lot of robotics companies have pivoted from consumer and entertainment to industry. We see companies in California making mistakes that we did years ago.”

In response to a question about whether it’s better to find a need or set up a feedback loop first, Jentoft said RightHand started “with a hammer” during the DARPA Autonomous Robotic Manipulation Challenge. “If you start with the hammer, you might throw away the handle,” he said. “It’s a very lossy approach. Some startups begin with technologies, but don’t chase nails.”

“You have to be flexible,” added MassRobotics’ Ryden. “You can’t be stuck on your own technology.”

3. A startup should stay focused, balanced

Crowe asked how each speaker balanced technical and business challenges.

“We see a lot of startups coming out of universities without understanding the business problems,” said Ryden. “They really need to spend more time with customers.”

“They also make the mistake of providing free demonstrations and trials,” he added. “If you have something of value, it’s OK to ask for revenue upfront.”

“Our three co-founders have Ph.D.s, but we signed up for a day as order pickers early on,” said Jentoft. “You need to get things to work with the business model, so you need balance.”

“We had a whole field of challenges,” said Cacioppo. “It’s important to prioritize them. We built a safe AMR [autonomous mobile robot], which is ‘table stakes,’ and then we layered on applications.”

“Engineers often want things to be perfect, but that doesn’t work in the real world,” observed Theobald. “You need to get something out there and then start a virtuous cycle of testing and getting customer feedback. Technical debt is OK if it’s a deliberate decision.”

Because Veo Robotics is focusing on safety, it can’t rush getting things out, observed Vu.

---

The Robot Report is launching the Healthcare Robotics Engineering Forum, which will be on Dec. 9-10 in Santa Clara, Calif. The conference and expo will focus on improving the design, development, and manufacture of next-generation healthcare robots. Learn more about the Healthcare Robotics Engineering Forum, and registration is now open.

---

4. Build a team, not just a collection of employees

“One advantage of the Boston ecosystem is that people like to solve problems,” said Jentoft. “Because RightHand Robotics is a hardware-enabled software play, we can’t have application developers turn over every six months.”

“The biggest challenge isn’t competing for customers; it’s competing for talent,” Vu said. “We don’t necessarily look for people with only robotics experience.”

“Companies are getting in earlier,” said Ryden. “Previously, Amazon would get college juniors to intern, and now, it’s reaching for freshmen.”

“It’s important to get people who like to work together,” said Cacioppo.

5. Fundraising is personal

“Fundraising is a personal question,” Theobald said in response to a question from Crowe about not relying on venture capital. “In 1999, I wrote a $5,000 check for Vecna Technologies, and we were entirely self-funded. It was a tough choice, but we didn’t want an artificial timeline for commercialization, and we waited for the market to be right. We also applied for SBIR [Small Business Innovation Research] funding from the government.”

On the other hand, venture capital funding can provide fast support, said MassRobotics’ Ryden. “Good VCs can bring customers, technologies, and a team,” he said. “It’s important to choose wisely.”

“When we raised money, we saw an industry that was moving fast,” said Veo’s Vu. “We love Boston, but there’s more interest here in funding biotech, so we had to go to California. It’s no coincidence that we found VCs such as Siemens in Silicon Valley with educational experience from Boston.”

“Lots of investors didn’t understand the platform or business model,” RightHand’s Jentoft said. “Find one that knows the industry you’re serving.”

“It’s exciting to have a high valuation, but it raises revenue expectations,” said Cacioppo. Shopify Inc.’s acquisition of 6 River Systems for $450 million closed two weeks ago.

“It’s also important to do your due diligence,” said Ryden. “A VC may have invested in a direct competitor and might just take a meeting to hear your story so it can go back and tell it to them.”

6. Pick the right partners

Going through a bad relationship with an investor or a partner is like a divorce, joked Vecna’s Theobald. At the same time, a good partnership or acquisition can benefit a robotics startup, said the panelists.

“We did not expect to sell the company, but at every meeting with Shopify, people wowed me,” said 6 River’s Cacioppo. “The founder is still the CEO and is still passionate.”

Corporate culture will change, noted the panelists, but it is always changing. For instance, iRobot’s initial public offering was fun and exciting, recalled Ryden and Vu, but relatively few companies go public. “Colin [Angle] instilled a positive culture, and he’s still CEO,” Ryden said.

Some relationships are international. RightHand Robotics this month opened its GK subsidiary in Japan. “Good communications are really important,” said Jentoft. “We have lots of early and late phone calls and visits, but we’re not only building trust in our technology, we’re also building trust in our company. Ask questions about your partner’s supply chain.”

7. Manage expectations

“Rodney Brooks once said that YouTube is the worst thing to happen to the robotics industry,” said Jentoft. “Customer expectations are shaped by science fiction. It’s important to know their problems and be aggressive in educating them. It’s also important to see robots in person in real time rather than in a show demo or an online video.”

“It’s a real problem when people don’t know the real value of robotics,” Vu said. “I’ve been building robots for 20 years, and fortunately, our customers are manufacturing engineers.”

An attendee asked about whether the companies represented had faced resistance from workers. Vecna’s Theobald said that many in the news media present “a false dichotomy of robots versus jobs. When we deployed at a FedEx facility, managers said to workers, ‘We’re automating because we want to keep you.'”

“Fear of the unknown dissipates as soon as people are shown how robots will make their lives easier,” he said. “Have a communications plan. Start interaction with workers early, and get their input.”

“We let companies pick their robots’ names,” added 6 Rivers’ Cacioppo. “It’s not enough for collaborative robots to be safe; they also have to feel safe, so that has to be part of the design. Our robots are built like a car, with headlights and taillights.”

“Robot behavior must be predictable so that people can form good mental models,” Vu said. “Veo has a monitor in the workcell that shows the operator what the robot sees in addition to status.”

“Understand what workers currently do, and allow for customization,” said Ryden.

A bright future for the robotics startup scene

Crowe asked about technologies that have improved or that the panelists would like to see improve more. Vu explained that gyroscopes were once large and expensive, but once Nintendo and others miniaturized them, more robotics applications became possible.

She also mentioned how batteries have limited robot and drone endurance and how prices for 3D sensing could drop from $50,000 to to $50. “In five years, there will be more flexibility in manufacturing, but lights-out is not the way,” Vu said. “We should be looking at robot interactions with humans.”

“There is a lot of value through pre-competitive collaboration. That way, we can grow the pie for everybody,” said Theobald.

“It’s an exciting time, as people begin to recognize where we can provide lots of value,” Jentoft said. “We’re starting to see robotics change the world.”