5 Solar Energy Breakthroughs with Machine Learning
Intro: Machine learning in the solar energy industry The high availability of data in the energy sector makes it a...
Traditionally, it has only been possible to look backward at the information collected from something in transit and guess the reason for delayed shipments, asserting blame to convenient scapegoats like “traffic,” when really the truck driver took a nap at a nearby gas station. Now modern tracking systems are bringing truth and clarity to the supply chain with real-time location updates, in-transit product monitoring, as well as delays and damage reports throughout the journey. Maybe the truck driver really needed that nap and that was the safest thing for her to do, now with IoT smart sensors, that can be tracked too.
Managing the supply chain pipeline means keeping a close eye on a multitude of processes. Reporting, dispatching, and tracking tools and staff, all need to work together to allow seamless production and delivery of products and assets.
The internet of things(IoT) or smart sensors are taking this challenge head-on. Offering new ways to measure and track previously un-quantifiable metrics. Combining these new data sources with advanced analytics is a strikingly affordable solution to solve a very expensive problem. By integrating multiple data sources, it is possible to gain a real-time understanding of the entire supply chain and to streamline the whole process.
From some context, a tech analyst company, IDC, predicts that in total there will be 41.6 billion connected IoT devices by 2025. While IHS, a global data and information services business, reports that by 2030, 125 billion connected devices will be part of daily life.
That’s why Krenar Komoni, CEO, of Tive, a leading IoT tracking device owner gives this advice:
“Trusted brands have to have a trusted logistics partner to give the end consumer the experience they expect. Top brands want to make sure that their product arrives on time, in good condition, and are available to their devoted customers.”-Tive
Ultimately, the ability to fulfill adapting consumer demands is the best metric for any successful supply chain and that be actualized through the adoption of IoT.
Leading supply chain companies have shown an interest in IoT solutions since they have repeatedly proven to have a direct impact on cost and productivity optimization strategies. Below are some of the most popular directions in which IoT solutions can impact the supply chain sector:
Quantifying these benefits depends on the use case and the industry segment. Let us take a closer look at some real IoT applications, backed up by companies who offer IoT solutions for the supply chain.
A variety of devices, vehicles, and expensive tools need to be maintained by supply chain managers. Their functional reliability ensures the stability of the entire pipeline and has a direct impact on operational costs.
Remote IoT sensors collect data from different pieces of equipment and enable preventative maintenance measures. Analytics performed on such data gives accurate predictions about equipment behavior and triggers notifications to maintenance staff before it causes delays in the supply chain.
ClearBlade’s CEO, Eric Simone, discusses a smart monitoring use case for an aircraft manufacturing company using IoT data to get real-time visibility into operational maintenance of their assets. ClearBlade’s software “ingests information about each assets’ location, elevation history, hours of use, etc. allowing optimization of maintenance schedules and compliance.”
Beyond knowing products are on their way via basic geolocation information, multi-sensory 5G trackers, like those from Tive, sense shock, moisture, and temperature for a comprehensive supply chain awareness, which are relevant for the most sensitive assets. Such is the case for medical supplies, fresh food, and other perishable products that are highly dependent on a maintained environment during transit.
Tive’s 5G ready sensors make it possible to constantly keep an eye on these products across an ocean transit or on the other side of the globe. From damage monitoring to proper handling supervision, they provide enhanced visibility for in-transit products. This, in turn, enables a better logistics experience for all parties involved and the minimization of associated risks.
“The system was able to alert a beer manufacturer that their shipment mistakenly went to 80 degrees Fahrenheit during shipment. The data from that incident allowed them to file an insurance claim to be compensated for their loss of product. The system was also able to tell a pharma distributor that their shipment went above 32 degrees Celsius when it was supposed to be between 2 and 8 C. In that case, they were able to alert the truck driver to correct the issue in real-time and save a few hundred thousand dollars on that load."- Tive CEO, Krenar Komoni
Awareness of carbon footprinting is a common characteristic of almost all companies in Gartner’s list of Chainnovators. While some have announced their intentions to actively try and reduce carbon emissions, some are planning to reach net-zero within the next decades.
Amazon is doing efforts in this direction and aims to reach net-zero carbon by 2040, 10 years ahead of the Paris Agreement. With their large number of physical assets and logistics infrastructure, they have already taken a series of steps towards this goal. Firstly, they have announced the acquisition of 100 000 electric vehicles that should start delivering in 2021. This shows commitment to the use of electric vehicles and is likely to set a global trend that more companies will follow.
An impactful example was presented by Garter as they recognized Western Digital for the ways their supply chain innovation was able to improve efficiency and reduce waste:
“Western Digital enabled cost savings as products are not in transit for as many days. Additionally, increased shipment consolidation means fewer shipments out of its facilities, leading to lower costs and fewer shipments on carbon-intensive planes, trucks, and ships, ultimately reducing its carbon footprint.”- Gartner
Elemica, Global Director of Marketing, David Cahn, mentions that they are “helping companies make green vendor choices by using electronic certificates of analysis during transit, which tell the source, quality, and all the details you need to know to make a sourcing decision.” Also noting that preferential treatment is given to green vendors, saying, “it is almost an imperative in certain industries that ingredients are sustainable and green.” Elemica offers a way to ensure those sourcing initiatives are enforced.
These examples show the trend of society to prioritize eco-friendly options to gain market share and the way IoT is used to aid green initiatives.
An IoT use case that spans well beyond the supply chain sector is that of safety and compliance. Safety measures concern all types of organizations that must ensure that they abide by the rules and regulations governing each phase of the product journey. From filing Chinese exportation documentation and ensuring expiration dates haven’t lapsed, to properly maintaining equipment to regulation standards.
IoT sensors can be embedded into machines but also attached to most any part of a system. A use case from ClearBlade utilizes data from a North American railroad company to ensure safety at railroad crossings.
“IoT sensor data informs railroad crossing systems of potential failures of safety measures like the lights, gates, and track switches. The system has the ability to see the amperage increase on a gate motor, and use edge-computing to know if this increase is enough to warrant a notification to be sent to alert maintenance staff about a potential malfunction.”- ClearBlade
Like this, they provide a real-time preview of their systems, which helps prevent hazardous situations before they occur.
Similarly, the value of asset tracking to avoid issues with regulatory compliance is evident in the following use case:
“An airplane customer had a critical asset go missing last week. This is a common occurrence that prior to deploying ClearBlade would have taken 2 people about 2-3 days to track the asset down. Typically when a new asset goes missing, a new one needs to be purchased to carry out critical aircraft maintenance in a timely fashion to avoid FAA violations and fines. In this case, it was a $20,000 asset and they found it in 30 minutes because of the ClearBlade system.”
Another example of compliance enforcement through IoT is from Samsara, whose AI-powered Dash Cam analyzes the road and generates real-time driver report cards for fleet managers if alerted to unsafe driver behavior. This enhances driver safety while reducing accident-related costs.
Today’s world is increasingly more connected through IoT-enabled machines big and small, which help humans to work smarter, not harder. By equipping machines with sensors, factory managers, logistics planners, energy grid optimizers, industrial farmers - you get the picture, can more accurately map machine workloads, inputs, and outputs. Ultimately leading to optimization and automation.
This input data continuously expands as technology is able to convert text(OCR), speech(NLP), photos, reviews, customer phone calls, and more, into measurable and quantifiable information. Machine learning is used to connect the dots and find previously hidden or misunderstood causal relationships and trends.
That’s where unified data lakes come in. They provide ONE singular place to put all this information and let the machine do its work at finding the hidden connections in the data. Fortunately, automation helps to ingest all these sources, by writing scripts and scrapers to go retrieve all this information from devices, social media, an old-school CRM, and while we’re at it, those old hard-drives from the ’90s. Dump that in there too.
As more data is available to apply to AI and ML algorithms the need increases for versatile integration and automation experts to garner real-time actionable insights.
Blue Orange Digital, a top-ranked AI co-development partner, speeds up enterprise IoT and Supply Chain companies’ innovation cycles, by assisting with the deployment of large scale custom analytics solutions. Using a data lake model, they can manipulate heterogeneous data types and gain data-insights in real-time. With the data available in a single repository, customized advanced analytics and visuals are made possible, such as anomaly detection and predictive forecasting.
“Blue Orange simplifies the process for IoT companies to make their data actionable. By building performant, advanced, secure infrastructure to enable teams to leverage real-time data for process automation, advanced optimization, and predictive analytics.”- Blue Orange Digital
“Firstly, we’ll feed these AI engines lots of data from IoT devices so they can create more advanced prediction models. Then you'll start to see a push of those models down to the edge, running them close to the machinery,” says ClearBlade CEO, Eric Simone.
Clearly edge computing at the device level will speed up the supply chain as it will ease some of the strain on the main data source and infrastructure. Instead, each IoT device will have the capability to run machine learning algorithms right in place, filter the noise, and only send critical information back to the data store.
Finally, Elemica had this to add, “IoT has taken us from descriptive to predictive, next is prescriptive. Meaning, devices will run logic algorithms and be able to react and adapt on their own in real-time. Like the sensor that sends an alert when a truck has exceeded its temperature tolerance but instead self-adjusts the thermostat itself. That’s what’s next.”
All these IoT solutions and the large amounts of data they collect have laid the perfect groundwork for continued advancements in predictive intelligence and advanced data analytics. Data aggregation and collection were the foremost obstacles that supply chain organizations faced but IoT devices have bridged that gap by digitizing most of the variables in the chain. These large scale real-time machine learning algorithms finally provide companies with the speed and accuracy necessary to implement proactive measures. Experienced teams have already had incredible success in implementing data infrastructures for a variety of supply chain solutions.
Special thanks to the following guest interviewees for sharing their perspectives on this ever-changing technology:
Know of an interesting Supply Chain innovation? Please mention it in the comments or get in touch, it could get featured in the next interview series.
Subscribe to get the latest on Machine Learning applications from Blue Orange Digital.