Last week I visited Emerson’s Chief Strategic Officer Peter Zornio for a closer look at the state of industrial automation and to learn how the internet of things is playing out for companies refining oil, designing chemicals and other process manufacturing. What sometimes gets lost in all of the excitement around the internet of things is that, for companies like Emerson or their clients, connected machines are not a new concept.
Their factories have had sensors built into the manufacturing process from day one. In the 90s, these sensors went from analog to digital and started reporting data to servers. But there were limits. The digital sensors still required wires, which meant that retrofitting wasn’t generally worth it, so manufacturers were limited to whatever sensors they had installed at the start.
A showroom shot of Emerson's automation setup.
But the cost-benefits equation changed with the ability to send data wirelessly and to power sensors with batteries. Manufacturers could add more sensors powered by batteries and start seeing more details.
The next big change, which is happening now, is that companies are aggregating sensor data to see when things might be about to fail, or to automate manual processes such as venting steam.
This is currently where much of the focus is for Emerson and its clients. Instead of taking direct measurements of processes, they put sensors all over the place and measure vibrations, environmental temperature and other elements to determine how these affect the process.
For example, factories used to measure corrosion by dropping a piece of metal in the vat and checking on it at prescribed intervals to see how it was faring. If it corroded, then the assumption was the vat was corroding as well, and should be replaced.
Now, other sensors located near the vat can offer a similar assessment without requiring direct contact and a person to manually check it. Other sensors might track vibration to see when something is about to break. In addition to preventing downtime, this can also prolong the life of a component that previously had to be replaced on a set schedule.
Zornio says the next phase of any industrial internet revolution will be taking the additional sensors and data they provide to optimize processes. For example, in the pharmaceuticals industry drugs are manufactured according to an exacting recipe. With more data, companies could adapt their manufacturing processes to the environment, something that can have a big effect when dealing with chemistry.
Of course, for this to happen, regulators need to approve how the drug-making process would be changed and companies will have to re-outfit their manufacturing plants. And with some process manufacturers still using 50 or even hundred-year-old plants, upgrades take time.
The other big change coming for any industrial plant is the ability to take data from a single factory, upload it to the cloud and then gain the ability to access the data from anywhere. This means that experts in certain processes no longer have to travel from manufacturing site to manufacturing site to troubleshoot or improve things. Instead of going to the data, it can come to them.
As an entire generation of experts in industrial manufacturing gets set to retire, this is no small concern for the oil industry and for other chemical manufacturers. Implementing more automation and more data gathering devices in their plants along with better analytics could help one person handle the job of several from a central location that functions almost like NASA's Mission Control does. In this case, instead of a distant telescope, mission control is monitoring a network of factories spread around the world. To take it a step further, those people may not all be employees of the company that owns the factory. They might be provided as part of a service contract by a company like Emerson.
Bringing this vision to fruition will require retrofitting factories, getting customers that are worried about data security on board, and partnerships with companies like Microsoft who provide the cloud computing back end where all of the data analytics take place. But Zornio is confident it will happen and he's confident that when it does, Emerson's ability to provide the expertise in data analytics associated with process manufacturing means its role in the industrial internet is secure.
Ayla Networks is the first agile IoT platform (Sponsored)
Ayla Networksprovides the industry’s first Agile IoT Platform, accelerating development, support, and ongoing enhancements of connected products for the Internet of Things. Ayla’s software fabric runs across devices, cloud, and apps to create secure connectivity, data analytics, and feature-rich customer experiences.
Offered as a cloud Platform-as-a-Service (PaaS), Ayla’s flexibility and modularity enables rapid changes to practically any type of device, cloud, and app environment. Learn more.
Startup Profile: Verdigris brings AI to energy management
Energy management is super hot right now...and not just because it's summer. GE spun out Current, a company trying to tie energy supply and demand together using sensors, weather data and predictive analytics. And Elon Musk, the CEO of Tesla is trying to take over SolarCity, a company providing solar panels.
Verdigris is a more modest effort, but its customers are clearly thinking ahead. Verdigris is a four-year-old startup from Mountain View, Calif. that makes a sensor that clamps onto a circuit inside your breaker box and measures how much electricity is used. Because the device clamps onto each circuit in a box, it can offer granular data about individual elements in a building, such as how long a particular light bulb has been on.
One of the custom-designed sensors that clamp onto a circuit. A system includes 40 of these clamps.
CEO: Mark Chung
Verdigris detects how much energy devices in a building use and applies intelligence to its findings.
Founded in 2012 and raised $9 million
Customers include Jabil, The Grand Hyatt, Autodesk
That data is taken from the device using the Verizon 4G network and analyzed in the cloud, where the information can be sent to a larger building management system or simply to a dashboard. The idea is that Verdigris' AI can help a building manager cut down on electricity use or look at the data to cut down on physical walk-throughs of a property. For example, in hotels the software and sensors combine to let the building managers know when a light bulb might be burned out, saving the staff the effort of checking.
Michael Chung, CEO of Verdigris, says one customer has tied the Verdigris artificial intelligence to a Siemens building management system. When the Verdigris AI detects the potential for a higher than normal electrical load, it can turn off some elements to cut down on power usage. This obviously won't work every time, but Chung says there are many elements that could be scheduled to run at alternative and potentially lower cost or lower impact times.
Verdigris has raised $9 million in venture funding so far and hopes to announce a new round soon. Chung hopes to expand the customer base to other clients, such as hospitals and manufacturing companies. There are a few companies trying to apply this level of granularity in residential homes such as Neurio and Curb, but in the commercial space the products tend to be less focused on such a granular picture.
Instead, startups like Building Robotics, which is applying AI to HVAC and Enlighted, which is trying to track energy usage through other sensors, are targeting different areas. Enlighted, Euclid and Digital Lumens have a broader mission to control everything in a commercial building while Building Robotics has a narrower focus.
From Chung's perspective, this means the Verdigris product can compliment several other existing building management startups. He's willing to work with them or the large building management companies such as Siemens or Honeywell.
So far the firm has over 150 systems installed in 40 to 50 buildings around the world.
wolfSSL is the industry leader in securing the IoT, as witnessed by our massive adoption across all IoT segments, including connected home, automotive, industrial, medical and smart grid. wolfSSL is used on all popular IoT operating systems and chipsets.
To support our widely adopted technology, we have a team of embedded SSL and cryptography experts available to ease implementation and assure the security of your device. We employ an open source business model, so you are welcome to download and review our source code. Check us out at www.wolfSSL.com/IoT
Insurance firms might be the savior of smart home technology. Because the price for many connected gadgets are so high, and consumers are uncertain if they are worth the investment, insurance discounts and programs are one way connected devices could find their way into a home. They also could help the insurance companies totally transform their business. This week, we discuss the future of the smart home and insurance with Ryan Rist, the VP of Innovation at American Family Insurance.
Before we get to that, Kevin and I talk about how manufacturers should kill connected devices, using the end of EyeFi as our case study. Then we offer consumers some advice on how to kill their accounts for connected devices when they want to return them to stores or just leave them behind, based on the experience a Redditor had with an Arlo camera from Netgear. And just for fun we also covered the Nest patent for a baby crib, the expansion of LoRa networks and my thoughts on the Arlo camera.
Hackster has completed a survey of more than 3,000 self-identified makers and concluded from it that about 71% of makers are hobbyists, 22% are pro and the rest fall into smaller categories. Among this population (which was 92% male) If This Then That was the top cloud provider, which makes sense given the high number of hobbyists, I suppose. IFTTT was followed by Microsoft's Azure and AWS. (For more on Azure's IoT skills see this previous newsletter.)
The survey also disclosed maker's favorite development platforms. Note that Intel's Edison isn't on this list.
News of the week
Cypress buys Broadcom's IoT business: After Avago bought chipmaker Broadcom, I was wondering what would happen to the emerging connected devices efforts Broadcom had built. IT didn't seem like a fit with Avago's networking and data center heavy focus. It wasn't. About 450 engineers, Broadcom's WICED IoT development kit and licenses for Wi-Fi and Bluetooth chips will all go to Cypress, which believes it has the ability to serve the myriad developer customers.
Meet Alibaba's smart car: Chinese internet giant, Alibaba has created a connected car that's available for pre-order (taking a page from Elon Musk, I suppose). Alibaba envisions the car becoming a link to services that it already offers, such as payments. But it also will tie into other internet services thanks to the inclusion of connectivity and Alibaba's YunOS. (TechCrunch)
In other auto news ... Intel, BMW and Mobileye, the company's whose AI is used in Tesla's cars, are working together to build a self-driving car by 2021. BMW's plans to actually deliver a car that doesn't require a human driver in 2021 could lead it to offer ride-sharing services. Essentially, like Emerson trying to turn its expertise in certain industrial processes into a service as mentioned above, BMW want's to turn its expertise about cars into a transportation service. Ford has a similar vision. (Fierce Wireless)
More worries about IoT security: Two security experts weigh in on the lack of security associated with current connected devices. Not a lot that's new here, but the story does point out the mismatch between appliance life cycles and the incentive to patch those products over their long lifetimes. If manufacturers can't find a solution, your 10-year-old connected washer could become a real liability. (ZDNet)
Why companies are avoiding the internet of things: Not everyone is jumping on the IoT bandwagon, according to data from the 451 Group. In the analyst firm's inaugural survey on the internet of things, it discovered why companies weren't deploying connected sensors and machines. The most popular reason? Worries about security followed by a lack of skilled employees. (Sensors Online)
Nest at some point considered a crib: Recode discovered a patent application from Alphabet's Nest subsidiary that shows a connected crib. The patent is from 2014 so don't get your hopes up. Yet, if I were going to make an expensive, highly designed connected device, a crib is probably a good place to start. Many Parents aren't price sensitive and they already are using the Nest Cam for a baby monitor. (Recode)
Comcast is betting big on the smart home: Not only did it announce the purchase of Icontrol's smart home management platform, but Comcast's Xfinity Home is designing its own line of connected devices. Xfinity Home is the cable giant's smart home platform. This week Comcast launched an outdoor security camera that will cost subscribers $199, or $99 if customers sign up for a $10-per-month video subscription plan. Comcast will sell the sensors itself and through Amazon. Comcast also works with a variety of third-party vendors of connected devices, such as Rachio and Chamberlain, and is emphasizing that its subscribers can expect Comcast to handle any challenges that arise in their newly connected homes. As Comcast worries about cord cutters cutting in on its average revenue per user, it looks like it thinks the smart home might help ease that loss. (Philadelphia Inquirer)
The case of the missing things: I think this story isn't actually asking what happens when things actually disappear, but rather when they disappear from our awareness, much like cords, light bulbs and the accouterments of electricity have "disappeared" for us. Either way, if you want to get all philosophical and provoke some thoughts about where the connected world and cheap data leads us, this is a good place to start. (Industrial IoT)
Are VCs searching aimlessly for the next big thing? Entrepreneur Elad Gil argues on his blog that this cycle of tech innovation is over and it's unclear what the next driver for high-growth startups will be. He views the move of venture firms into food, manufacturing and even physical goods as an indication that we have reached the end of a tech cycle. (Elad Gil)
More than you ever wanted to know about traffic lights: One of the biggest promises associated with smart cities is better traffic flow. This deep dive into a Dutch town's traffic lights shows how it can be done. (Bicycle Dutch)