What is IoT and Machine Learning?
Imagine the capabilities of your smartphone melded into every imaginable component. The Internet of Things describes a vision of the future as much as it implies a set of underlying technologies. It envisions a situation where everything that should be connected is connected. This currently includes Ring Doorbells, ‘smart’ appliances, and Apple’s seamless connections between watch, iPhone, and AirPods.
Where IoT describes connected devices, Machine Learning is the computer intelligence that draws insights from and controls these IoT devices. ML involves training a model, then using that model to draw conclusions from future data. ML algorithms are the methods with which ML models are created, similar to how different cooking methods exist to cook the same food. ML technology brings us self-driving cars, facial recognition, and fraud detection.
Two main synergies exist between these technologies. First, connected devices provide better plentiful data, allowing for more accurate models. Second, IoT devices allow more possibilities in which ML intelligence can be applied.
I had the chance to (virtually) sit in on Claris’ Engage 2020 panel dedicated to present and future implementations of ‘Internet of Things’ enabled devices and Machine Learning.
Two Awesome Examples
Eric Astor – Furnace Records
Eric shared Furnace Record’s transformation through integrating IoT sensors into his production process. Eric has 12 years of experience producing custom records at his factory and now has new levels of precision in his production process. During the retrofit of his factory, he added PLC / IoT devices into key dials, pipes, and valves throughout his plant. This IoT network gathers the data on all input factors automatically and finely tracks them from one central panel.
With this gathered data, Eric and company are able to analyze past results for production efficiency insights, and also conduct predictive equipment maintenance during factory off-hours, resulting in less factory downtime. He has achieved greater efficiency and quality control – freeing his team to focus on core creative processes.
Gery Pollet – Blyott
Gery described the challenges and motivations in founding Blyott, his startup focused on IoT for location-based tracking and monitoring. A hospital had come to him looking to track equipment throughout their building in real time. There weren’t pre-existing alternatives, so he wound up researching and engineering the solution from scratch. Among the many technologies explored were RFID, wifi, Bluetooth, and Ultra-wideband.
A typical 500 bed hospital tracking 10,000 pieces of equipment will gather at least 60,000 points of data every minute. The scale of this data lead him to settle on a software architecture using Amazon Kinesis, DynamoDB, and AuroraDB culminating in the final FileMaker presentation layer.
With this system, Hospital employees are able to locate any piece of equipment throughout the hospital, simplifying item retrieval, inventory purchasing, and minimizing equipment loss.
Takeaways & Considerations
Additionally, panelists Niousha Gouda (Apple), Steve Winter (Matatiro Solution), and host Ross Rubin (ZDnet & Reticle Research) rounded out discussion on IoT/ML applications.
Niousha shared that applying AI in business allows many opportunities for differentiation, and that cloud infrastructure and personally trained ML models make this increasingly more accessible. The many different fields of AI lend themselves to all types of business applications: Automation of repetitive tasks, Data & Sensor Driven AI, and NLP.
Gery fielded a question regarding security of IoT devices presenting ‘5 billion’ vectors of attack. His rule of thumb was to ensure each point is operating on end-to-end encryption and make use of certificates. The entire infrastructure must be built in a very secure way.
In making the transition to greater levels of digital connectedness, we hope to apply wisdom and curiosity in creating greater utility. At Codence, we regularly navigate the tides of technology with personal and professional interest.
Are you already implementing IoT and Machine Learning in your business flow? If you have any ideas you want to bring to life, reach out to us.
Information Specialist / Apprentice
Truman is joining us as an apprentice to learn FileMaker to transition from this apprenticeship to a support role for one of our clients. Truman worked as a SysAdmin before getting into 42 Silicon Valley college. He is interested in civic participation and locally sourced foods.