Driving Sustainability with Connected Intelligence
One way to take action is through the Moab Foundation, a non-profit organization working to create a more sustainable planet through the application of connected intelligence.
Wienke Giezeman, CEO and Co-Founder @ The Things Industries, moderates this panel of IoT leaders to find out what went wrong and where we go from here. Panelists include:
It may or may not CONNECT to a remote monitoring service over the Internet.
It COLLECTS air.
It ANALYZES the air looking for smoke using either an ionization chamber or photo-detector.
It ACTS on the insights derived from this analysis and sounds an alarm if smoke is detected. The battery on this device typically lasts a year before the annoying beeping sound begins.
Amazingly, this self-contained IoT device and analytics platform doesn’t require mythical AI powers to deliver value to the customer.
Don’t overthink it.
We weren’t at the MIT Media Lab this time around, so we did a virtual WFH Edition. I’ll delivered a keynote on how #DigitalTwins live at the intersection of #IoT and #AI. I hope you enjoy it.
Watch my presentation and others at:
https://www.verypossible.com/resources/video/mitef-connected-things-wfh-edition
With the ability to deliver the state of machines and analyze that information in near real-time, the Internet of Things facilitates new business models beyond typical CAPEX purchases.
It was an insightful event with amazing audience engagement. We had a candid conversation around the challenges we all face in IoT projects and how we can get better. What we have learned to date and what has gone wrong and well?
Our distinguished panel included:
IoT Successes and Challenges: Commercial and Industrial Uses
This means following the same strategy you use when taking a big test. Do the easy stuff first and leave the hard things for later. So how does that apply to connecting your organization’s people and things to drive additional value? Read on as I illustrate some simple examples to get you started this year:
People doing remote work in the field still capture information with a paper and pencil. Then they drive back to the office at the end of the workday to transcribe their scribbled data into their organization’s back office computer system along with a dose of human error. I know you’re thinking this is impossible in the 21st century but I see it all the time. Thanks to the magic of cellular data networks combined with smartphones, tablets and smart, connected IoT devices, this inefficient activity can come to an end.
If the remote activity requires dynamic, person-to-person interaction, data should be captured and validated by a smartphone or tablet app and wirelessly transmitted back to the organization. If the remote activity is an inspection that consists of taking an analog reading, drop the clipboard and utilize one or more sensors to convert analog values to digital equivalents and wirelessly transmit the data back to the organization. While newer machines may include the built-in compute, networking, storage and sensors to get the job done, most of the world is filled with older machines that must be retrofitted with these capabilities. If retrofitting isn’t possible, then visual inspection of an asset can be conducted by a fixed or mobile camera where the photo transmitted back to the organization and computer vision converts the analog image to digital values.
This is the simplest first step you can take in augmenting and/or replacing an expensive activity with connected intelligence. Without employing analytics, you cut costs by reducing or eliminating remote activities that include travel, vehicles, and fuel expenses just to name a few. You lower your risk and liability by reducing the need to put people in vehicles or having them perform inspections in precarious or otherwise dangerous situations. You gain speed and agility through the instant availability of data allowing your organization to respond to problems and opportunities more quickly.
While we’re all excited about what the future holds, it’s important to interface with the world as it exists in the present. The overwhelming majority of the Things that are all around us have existed for years or even decades. In order to derive valuable insights from these operational technology (OT) machines and environmental systems, you’ll have to connect to them in often unfamiliar ways and learn how to speak their language. Achieving broad success requires you to be comfortable with brownfield projects. In many scenarios you’ll find yourself using low-speed serial cables and communicating via 100+ wire protocols and data formats.
You better have your Rosetta Stone handy. Don’t be surprised if you must interface with a programmable logic controller (PLC) instead of connecting to a machine directly. Oftentimes, the OT folks in factories won’t let you get near their machines. If you want to be successful, you’ll have to make peace with this OT/IT reality.
Many of these existing systems are either unconnected or connected to a closed system often built by the same manufacturer. Imagine dozens of machines on the shop floor individually connected to their own, proprietary analytics systems. By freeing the data found in these systems, you go from having countless data silos to a achieving a blend of machine, environmental, organizational and 3rd party data. This delivers much needed context and allows you to see the big picture across systems of systems to make better decisions.
You know how all the industry analysts throw darts at a board and tell you about the tens of billions of connected devices we’ll have in the coming years? They keep having to backtrack on these predictions because of one very important reason. Bringing the Internet of Things to life is still largely a very manual process of configuring devices, networks, code, platforms and security. IoT today operates like one big aftermarket car stereo store. For the Internet of Things to be the huge success we all want it to be, we have to remove many of these manual processes and leave custom work to those targeting specific use cases.
It must all start when Things are created on an assembly line by original equipment manufacturers (OEMs). The smart, connected machine must have compute, storage, networking, sensors and actuators baked-in from the very beginning. In other words, a microcontroller, cellular module, software, a trusted platform module (TPM) and associated security tokens or certificates from the get-go. This new product will be created in one country, possibly shipped to a distributor in another country, and purchased by a customer in yet another country.
When this smart, connected product wakes up somewhere in the world, it must first make an automatic connection to a local mobile network operator via its cellular module and associated connection management capabilities. Next, it must use that connectivity to access the OEM’s globally-available service and pass along its identity and security credentials. This service will determine what and where the product is, who bought it, and ultimately what IoT platform it should send its telemetry to and receive commands from.
By following this process, much of the friction that’s holding the Internet of Things back is removed. Time can be better spent targeting specific use cases allowing customers get to value more quickly at a lower cost. This better use of human and machine resources will exponentially accelerate the rising tide of IoT that lifts all boats.
I know you’ve heard a lot about Edge Computing over the last several years. As it relates to the Internet of Things, the Edge just means moving compute and associated data filtering, aggregation and analytics closer to the machines and environmental systems that actually create the data. When the IoT megatrend heated-up ten or so years ago, companies that weren’t familiar with the decades of capturing telemetry and controlling systems via M2M and SCADA systems assumed IoT data must go to one of the many public clouds. Unacceptable latency, high broadband costs and data governance issues gave rise to concepts like the Fog and Edge to mitigate cloud shortcomings. Since I’ve spent most of my time in the industrial space, this has typically meant placing edge compute near machines on the factory floor or even on a bullet trains where decisions could be made in milliseconds. Performing computing tasks at the Edge has also worked well for discrete and process manufacturers who say, “the data doesn’t leave my factory.”
More recently, the telecom industry has thrown their hat into the ring by placing distributed compute infrastructure and resources at the edge of service provider networks where “last mile” content and applications are delivered. While this particular Edge isn’t on-prem, the concepts of supporting low latency, data intensive applications still applies since it runs at the edge of cellular networks and is significantly closer to the source of IoT data than distant public clouds. It also provides the benefit of reducing congestion and signal load on the core network, so applications and analytics perform better. This architecture is sometimes referred to by the acronym MEC which can either mean mobile edge computing or multi-access edge computing. Expect to see this type of Edge compute used heavily in smart cities, public infrastructure, faster video games with reduced ping times and vehicle-to-everything (V2X) scenarios.
Moving compute resources to the Edge benefits myriad IoT use cases including split-second application responsiveness, reduction in bandwidth costs and congestion, plus the granular data governance needed to meet local, city, state/province, and country security + privacy requirements. Public clouds still have their strengths. As always, just use the right tool for the job.
When you don’t put a person in a car, truck or plane to perform a remote inspection, you’re not burning fuel or congesting freeways. Think of wireless data networks as your replacement for travel. Using smart, connected, new machines and retrofitting older machines ensures you always know about their health and performance so they can operate more cleanly and efficiently. Edge computing allows you to address problems more quickly while alleviating network congestion. As always, think of the Internet of Things as your early warning systemto detect pollution, fires, water leakage, unsafe machinery, excessive electricity usage, deforestation, water contamination, and thousands of other important use cases.
As you can see, none of the topics I discussed required Machine Learning, Deep Learning, Neural Networks, or any kind of Artificial Intelligence to move the Value Needle. Throughout 2020 I want you to avoid the hype that bombards us from every direction and focus on specific problems that can be solved in your organization through the use of Connected Intelligence. Steer clear of esoteric technologies and concepts that you and your colleagues struggle to wrap your head around.
If you start small, keep things simple, and iterate steadily throughout the year, I know you can knock it out of the park and derive tremendous value for your organization while being sustainable.