For the tech pioneer Gabe Glynn, factory safety is a very personal matter
Factory safety is not a theoretical issue for Gabe Glynn, CEO of the wearable tech firm MākuSafe. “The reality is I wouldn’t exist today because of an industrial accident—had it not been for my great grandfather walking out of a factory in Cedar Rapids 101 years ago, just moments before it exploded,” he said in reference to John Griffin. “He spent the next few weeks pulling the bodies of 43 coworkers out of the rubble.”
Today, more than 1,000 people around the world die each day because of work-related accidents, according to the International Labor Organization.
“People read articles about us and call us, and they say, ‘Hey, my brother died in a work accident. My family is praying for you guys’,” Glynn said. “That stuff gives me chills. The reality is, here we are 101 years after that explosion in Cedar Rapids, and too many people aren’t going home from work. We are significantly better. But we still can be better yet.”
To that end, MākuSafe began offering its safety solution for sale in March—with about one-third of its available inventory for the year committed to a variety of industrial end-users and workers-compensation insurers, such as EMC and AF Group.
The firm uses a subscription model. A one-year subscription, which includes all hardware, software data access, installation and support costs $22 per wearable each month. One device can be shared by a first-shifter and a second-shifter but it must be recharged during the third shift. So, a company that subscribes to 20 wearables to cover 40 employees would pay $5,280 a year.
The company has some indirect competition, including Corvex and Guardhat. But they are seemingly focused on construction and oil and gas, respectively, Glynn said.
While Glynn’s family history informs MākuSafe’s raison d’être, the idea for the company’s product came equally from podcast interviews Glynn undertook after he sold his first company in 2015. Flush with a bit of cash from that sale, he bought some old audio equipment, watched some YouTube videos on how to create a podcast and set out on the road to relay to the world what life was like for manufacturers here in Iowa.
In addition to the background of his great grandfather, Glynn’s father was a machinist and is currently a factory safety manager. So he expected there was enough fodder for a successful string of manufacturing biz interviews. But he did not expect that the harrowing stories he heard as the founder of the Advanced Manufacturing Podcast—one factory worker died of heat exhaustion; another lost his hearing—would give rise to a new company focused on helping manufacturers avoid catastrophic losses.
Glynn founded MākuSafe four years ago with his friend and IoT and cloud-computing specialist Mark Frederick, and the wearable tech company’s first product is already showing promise inside factories.
The wearable, which took three years to develop and is manufactured by NationGate in Penang, Malaysia, packs communication technology and MākuSafe sensors for light, sound, temperature, humidity and air quality into a 1.5 x 2.5-inch device.
“It also has an accelerometer for tracking motion,” Glynn said. “We look for things like slips, trips, falls, and forceful repetitive movements. And it has BLE (Bluetooth low energy) onboard, and wi-fi,” which help with location positioning.
The sensors regularly monitor environmental conditions and capture any that are outside normal range, in addition to signatures of forceful human motion, and relay the data to the “MākuSmart” cloud dashboard within 45 seconds.
MākuSafe then employs artificial intelligence (AI) and machine learning to spot dangerous trends and likely hazards—and moves to alert plant safety leaders right away, increasing the chances that accidents can be avoided.
Additionally, the device includes a button that lets users record a voice memo on the fly about near misses or good-catch observations that the sensors might not pick up.
“If a bunch of pallets falls and lands next to me, the device isn’t going to know that. But it could have killed me,” Glynn said.
The device is also capable of integrating with automation systems.
“We envision a near future where our customers could rely on MākuSmart to automate risk out of the environment based on trends in the data gathered from their facility,” he said.
Safety is a ‘living, breathing thing’
Glynn’s podcast interviews revealed a serious need for greater factory safety, Glynn said: “We should be constantly monitoring environmental conditions because it is organic: Every day, it’s different, Every hour, it’s different. It’s a living, breathing thing.”
On top of that, putting sensors on walls is likely to miss what’s truly happening, he asserted: “We said, ‘Let’s take these stationary sensors people are putting on walls, put them all together and put them directly on the person—because the environment can change in a matter of feet.’
“If you’re on one side of the machine and I’m on the other, the sound is totally different. And the air quality can be different,” he added. “So, we said, ‘Let’s create something that looks outward at the sphere of environment around the person’.”
Now that mobile sensors can, with AI and machine learning, detect the first inkling of a developing danger, it makes little sense to operate a factory without them, Glynn asserted: “Think of it like dominoes. When something bad happens, we can see which domino was dropping before that—and which one dropped before that. Eventually, we can get to a place where we’re seeing so far out that when a domino begins to drop, we already know what the result is going to be” and can stop the accident or exposure from happening.
When Glynn was growing up in the 1980s, he experienced out-of-the-blue tornado sirens that gave him and his family “a few seconds to get in the basement,” he said. “Today, it’s like Monday morning and the weatherman says, ‘Keep an eye on the sky Saturday afternoon from 4:00 to 6:00.’ And he’s right most of the time.”
He and Frederick, the company’s CTO, reasoned that, with the right data gathered, they should be able to provide manufacturers with similar, advanced warning of impending trouble.
And, Glynn added, “we’ve been able to identify where employees are being exposed to sound levels that are potentially harmful for extended periods of time. This gives a safety leader useful intelligence to adjust their hearing conservation programs and keep their people safe.” OSHA established permissible exposure limits, such as an average of 85dB for an eight-hour period. But if the dB level rises to 90, the time period is cut in half.
In testing of MākuSafe’s first device, it also picked up “really high-impact, repetitive motion” in an industrial rug-cleaning operation, he said. “The employees were jerking the material out of the shoot to roll it up and move on to the next one. That is going to lead to shoulder injuries. That is going to lead to losses.”
MākuSafe also tracks total volatile organic compounds and CO2—using “a pretty low-cost sensor,” Glynn said. “It’s not a $2,000, highly calibrated CO2 sensor. But it’s designed to identify clues.
“We were testing in a facility, and one day we saw that all the employees had really high levels of CO2. So, we contacted the insurance company, which sent out an industrial hygienist”—who discovered that a machine was ejecting a detergent into the air that caused the MākuSafe sensor to pick up what it thought was CO2.
The air quality problem was solved by simple maintenance on the machine.
Glynn has a long list of examples like that, which is part of the reason insurance companies around the world are paying close attention to MākuSafe. In November, the 12-employee startup won the 2019 ACORD Insurtech Innovation Challenge.
Startup raises $6 million
Ramco Innovations, a supplier of control system builds, robotics and sensor technology, was one of the first investors in MākuSafe, which has since raised more than $6 million in venture capital.
“The draw was they had an exciting and innovative emerging technology that really captured a market related to safety,” Ramco CEO Hank Norem said. “So, we were excited to be part of their journey.”
Norem and Glynn shared a stage at a tech conference about six years ago. Norem became Glynn’s mentor. And when Glynn visited Ramco’s prefab offices and spoke about the concept for MākuSafe, Norem offered an old copy-machine room as MākuSafe’s HQ. Today, in that space and up in the attic, MākuSafe has folks testing software and the user interface.
Norem also set up Maple Ventures, a business incubator that has so far cycled through two other startups. The participants share a CFO, collaborative workspace, marketing resources, warehousing and engineering services and IT functionality.
The arrangement gives Ramco a first look at emerging technology, Norem said.
“Even if a company doesn’t end up in Maple Ventures, they come through and explore the space. At the same time, there are other related parties working with folks like MākuSafe that get a peek inside the walls of Ramco and check out our automation and robotics center and look at our capabilities. So we get introduced to companies through other companies that are working with these startups. The other thing is, it’s a potential partner for us at some point—whether it’s a customer or supplier relationship on the distribution side of our business.”
And it has proven invaluable for MākuSafe, Glynn said: “In the early days, when we were testing our technology, we would have to drive 45 minutes to a factory we were working with to test one thing out, and then drive 45 minutes back home. Being in an industrial space has allowed us really, really quick iteration,” as well as valuable warehousing space.
Crowdsourcing on the horizon
MākuSafe’s first production devices arrived from Malaysia in June. They all went into manufacturing environments for a final round of paid pilots in six states. The sites included light-duty manufacturing and a steel foundry. Some were coordinated with insurers; others were negotiated directly with industrial organizations.
Glynn and Frederick are providing them with very specific remediation actions “so that they can go out and have a meaningful conversation with their employees,” Glynn said. “Something as simple as, ‘Go out and do a 30-minute observation. See why the employee is jerking two G’s of force over and over again.’ We curate a bunch of different content from different sources. Their insurance company has training videos, for example. We can also bring into the platform trainings from the National Safety Council, state organizations, whitepapers and anything else that may be helpful at the moment of need.”
And they follow up to see what impact the remediation actions have.
But they are chomping at the bit to expand to the point where they can crowdsource the best remediation actions, Glynn said. “Our system will get smarter and know, ‘Alright, this problem is arising for this new person down the street—but we already know how to solve the problem: Here are the best remediation actions’.”
In pilot programs, 2 manufacturers take in a ton of new information
Capturing near-miss data at Quality Manufacturing
In testing that began in August, the sensor-filled, arm-worn safety device from MākuSafe has demonstrated the instantaneous, automated recording of slips, trips and falls at Quality Manufacturing in Urbandale, Iowa, said Sam Huffman, safety engineer.
EMC, an investor in MākuSafe and the company that provides the steel parts maker workers’ compensation insurance, encouraged a pilot of MākuSafe’s device. The pilot began in late August, with 20 units in a single building. It took less than a week to map out and install everything and explain it all to employees.
“Once the employees understood that incidents, like slips, trips, and falls, were being mapped as opposed to their movements, they agreed to wear the devices with no problem,” he said.
By mid-October, the device had picked up a few slips, trips and falls, Huffman said.
“To us, the most interesting piece of the MākuSafe device is capturing near-miss information. It’s easy for an employee to not report something like that” or to forget it happened by the end of the day, he said. “Now the employee can just press a button on this device and make a voice recording of what happened—and whoever is monitoring the MākuSafe software gets that fresh data, including when and where it happened. It lets us immediately look at the situation” and avoid future injuries. “We’ve had employees do that.”
In addition to protecting factory workers, companies can use MākuSafe’s device to reduce the number of insurance claims it must file and therefore reduce worker comp costs, Huffman said.
MākuSafe leaves no stone unturned: It monitors environmental conditions, such as noise exposure, light exposure, air quality, temperature and humidity, as well as potentially hazardous human motion. It maps the location of these occurrences. And it enables voice reporting of near misses and observations from the front lines, he said.
All of that data can help companies effectively deal with OSHA and buy the right personal protective equipment, such as safety glasses and earplugs, he added.
Experiencing machine learning at Crest Foods
MākuSafe’s armband revealed in a pilot at Crest Foods that it can adapt to new environments quickly via machine learning, said Karen Yardley, safety manager at the dry foods packaging plant in Ashton, Illinois.
In coordination with the insurance firm United Heartland, 750-employee Crest Foods has been using 20 MākuSafe devices since mid-October. It served as the company’s foray into wearables.
“The most interesting thing is seeing how the technology interprets some of the actions out on the floor,” Yardley said. “A ‘slip, trip or fall’ may be some other exaggerated motion with your upper body,” such as swinging open an elevator door.
The MākuSafe device can be programmed so the door-swinging doesn’t continue to register as a problem, she said. “It is learning about all of the different actions we do here at Crest. Today, for example, one of my employees said, ‘Oh, I wanted to let you know I accidentally popped the device out of the armband and it fell and hit the floor.’ I pulled up the dashboard and looked at her activity, and sure enough, you could see on the graph that it happened—but the MākuSmart cloud software had already sorted it out and said, ‘We know that that wasn’t an accident or injury’.”
The safety device might “really help us be much more proactive, instead of waiting for someone to get hurt,” Yardley said. “I now have a short list of things I’m planning to take to my safety team for review: Things we may do day in and day out,” such as using a plastic mallet to break up a heavy, dense product, but that the safety device called out as in need of risk assessment.
How to get started with a wearable
MākuSafe conducts a site visit, to gather some intake details and preconfigure software, and in about half a day can complete the following steps to get a company up and running with its armband device:
- Place and train a small number of Bluetooth beacons throughout the plant, which are used to identify location;
- Install a wall-mounted base station, which is used to charge the wearable devices and also serves as the hub for communicating with MākuSmart cloud software;
- Provide brief orientation training for armband wearers, as well as setting up safety leaders with access to the data being collected;
- Run quick trials to confirm connectivity, location identification, and real-time sending of data and voice recorded memos to the cloud dashboard, and
- Ensure leaders can use the micro training resources provided in the platform for ongoing communication about: how workers check devices in and out for every shift; how to wear the armband properly, and how to report near misses.