The Device Chronicle interviews Dr. Tom Bradicich, HPE Fellow, Hybrid Cloud, AI, & Industry Solutions.
Introduction: A.I. on the Edge in Building Management Systems
Tom speaks on the IT / OT convergence and AI on the edge and building management systems. We asked Tom to share his perspectives on this timely topic.
Recently, Northern Tech CTO and Mender co-founder Eystein Måløy Stenberg identified three key trends in building management systems:
- Climate tech transforms buildings into smart distributed infrastructure
- Remote management and AI drive cost savings
- National cybersecurity and supply chain concerns drive open solutions.
In this interview, Tom identified and shared his thoughts on a 4th trend:
Operational technologies converging with IT in building management systems.
Chapter 1: Implications of convergence between OT & IT
Tom begins the interview by stressing that the control and actuation part of OT is essential. It's the ultimate reason for a building management system's existence and is predominantly related to an IoT deployment. You want to control the things: doors, windows, gates in the building, or the cameras, the HVAC, the air conditioning systems, the heating, lighting, the utilities and plumbing, robotics, and automation that go on within the building. You also want to control the access that people have within those buildings. A badge reader or an authentication system could be the point of control.
3 PCs
When you walk out of the "data center" with the IT capabilities, you get to the "edge." Buildings like wind farms or power stations are at the "edge," and three critical dimensions of an edge in IoT deployment relate. Tom describes them as the three PCs.
- Perpetual Connectivity: Everything is connected perpetually effectively.
- Pervasive Computing: The computing capabilities down to a device such as a smartphone are increasing tremendously
- Precision Control: information to help you control various things, a pivotal aspect of how you would build products and how you would help customers achieve solutions.
Three convergence types
Now, blend IT (computing, computers, storage, and networking - usually Wi-fi or Ethernet) with OT (controls, data acquisition – DAQ, sensors and actuators), and Tom sees three types of convergence.
- Physical: Using the smartphone metaphor, all the critical functions come together in one self-contained unit.
- Data: OT data such as temperature and vibration can be fed into IT, such as financial systems, to facilitate the prediction of maintenance and replacement costs.
- Process: Those in the physical space, such as Line of Business facilities managers in a building from the line of business, are working closely and converging with the CIO in charge of protecting the IT mothership.
Chapter 2: A.I. in loss control in retail & video surveillance in intelligent buildings
A.I.-ification
Tom believes that whereas electrification and data-ification were demanded by the masses, "AI-ification" is currently being pushed. AI has great potential. Tom gets enamored by applying AI analysis and analytics to a video. He remarks, "If a picture is worth a thousand words, then a video is worth a million. And a video with A.I. analytics is worth a billion words. And therefore the power of video and the ability to analyze what's happening with it."
Two problems are solved by applying AI to video in a building space:
- Anticipate and prevent nefarious acts such as theft by applying AI and analytics to video footage.
- Goods depletion rates can be surveyed, analyzed, and replenished quickly at peak times of demand, such as a holiday season or a major sporting event.
Chapter 3: Access control & A.I. surveillance reduces operating costs & increases security
AI has tremendous potential to increase customer satisfaction when combined with facial recognition and analytics. Tom and his team created a prototype candy vending machine that only dispenses candy when the person using the device smiles. The cyber-physical machine can respond to the person's mood and behavior. Mood predicts behavior, so this can have powerful applications. Tom also worked with a large airline on detecting VIP customers waiting in a counter cue. The airline attendant can go out and greet the VIP customers and provide them with immediate, high-quality customer service. The experience is rapid and pleasing as all the essential VIP customer information is available to the attendant.
Shift left from data center to edge
Tom sees that all the data center's power is shifting left to the edge. Tom shares the example of the HPE EdgeLine 8000 Series that he and his team developed, which delivers "enterprise-class computing with tremendous amounts of storage" to the edge. But still, Tom sees some business challenges with the "edge," mainly around the security and manageability of the devices.
Chapter 4: Best strategy for updating & securing the software on edge systems
Security inheritance
Chapter 5: Security Inheritance Concept for OT & IT systems networked with entire software stacks
To conclude, Tom presents the concept of "security inheritance" in the OT / IT convergence. OT functions such as control systems and data acquisition systems can enjoy security benefits from conventional IT. Tom envisions a control system and an actuation subsystem now physically placed in the same box and frames and covers as a server and computer. It is physically in there, electrically and logically connected. You get immediate performance benefits. Such integration and convergence affords “management inheritance”, too. Manageability is being able to upgrade, patch, and so on. You can implement these best practices and tools into OT - control, data acquisition, industrial networks, and protocols.
Key takeaways on A.I. from Intel Innovation 2023
Tom also attended the recent Intel Innovation Conference. He had some great takeaways on A.I. to share from this seminal industry event. Watch this video for more on that.
Chapter 6: Key Takeaways from Intel Innovation 2023
Please contact Tom at https://www.linkedin.com/in/tombradicichphd, and he'll be happy to share more perspectives on his segmentation of the AI marketplace, the 12 challenges of the edge, and the 7 fundamental principles for computing at the edge.
Follow Tom on Twitter at twitter.com/tombradicichphd and HPE at twitter.com/HPE.