On Thursday, Jan. 31, Ontario’s Research Innovation and Commercialization Centre ( RIC Centre ) presented the first of six workshops in Mississauga, Ont., as part of its aptly named Driving Dreams series. Making a strong first impression, it’s clear that this workshop was catered toward those who are driving forces in industry as it provided an opportunity for networking, collaborating and resourcing.
The event brought together industry leaders from automation, security, Internet of Things (IoT), machine learning, hardware and more to discuss the theme: Industry Trends and Market Opportunities – Sensors, Miniaturization, Technical Convergence.
“The RIC Centre is one of the incubators within the Toronto marketplace,” said Trevor Bingham, Director of Business Development and IoT at Arrow Electronics. “It differentiates itself by helping customers speed their time-to-market by providing in-house resources.”
Bingham also added that one of the RIC Centre’s strengths is “leveraging partners within the community to help their clients speed up time-to-market while providing solutions in areas such as commercialization and export.”
The day began with Bingham introducing Arrow Electronics’ Business Development Manager, Michael Willis. Discussing some of Intel’s recent and upcoming product lines and portfolios, Willis held a hands-on showcasing. Before his first slide, he encouraged attendees to throw their business cards into a raffle to win one of the few Intel Neural Compute Sticks he had brought with him among other goodies. The grand prize was one of Intel’s NUC Mini PCs.
Afterwards, attendees were invited to sit down with some of the experts present in a conference room where they could hold a techstorm and discuss ideas and solutions to bottlenecks in the development cycles of their own projects.
Alex Fesiak, IoT Business Development Manager at Arrow Electronics, was the keynote speaker. His talk focused on his experiences, opinions and insights into CES 2019 this year. He brought up five main sectors to keep an eye on:
1. 5G and interconnectivity
2. XR: Augmented and Virtual Reality
3. Automotive technologies: autonomy and safety
4. Privacy, ethics, security and liability
5. Artificial Intelligence
“Many of the things that we saw at CES 2019 were the result of many customers that we are affiliated with who embed their technologies in 5G or into processing AI,” Fesiak said. “It always resonates Arrow’s Five Years Out with me because we truly have a handle on what is going to be happening in the future and where some of these technologies are going to lead us.”
Following Fesiak’s talk, a panel began which was moderated by RIC Centre’s own entrepreneur-in-residence, Paul Barter. Joining them were Dan Mathers, president and chief executive officer of Eleven-x Inc., and Feisal Hurzook, chief technology officer at Archronix Corp.
Questions from the crowd included an inquiry into Ontario’s transition from hardware to software and its global implications (Mathers touted Marc Andreessen’s saying, “software is eating the world,” but also defended Ottawa’s history as Silicon Valley North). A question about security in IoT, IIoT and autonomous vehicles allowed the panel to discuss opportunities for new products and services.
“Some of the interest that I heard today was electric vehicles,” Hurzook shared. “It was very interesting: solar power, batteries, AI, connectivity of floor assets, and more.”
Internet of Things: Connectivity and Miniaturization
Hurzook explained that one of the main goals of IoT is “saving the world from inaccessible data,” by asking hardware engineers the question: “How do we get the data to the Internet, so they can be analyzed in the cloud and be used?”
Hurzook wasn’t alone in stressing the importance of collecting and analyzing the increasing amounts of data generated as a result of IoT. Similarly, Willis explained that the key is optimizing how analytics are done at the edge .
“More computation and analytics are happening at the edge,” said Willis. “That is, where the data is being generated from a sensor stand-point—whether it is a smart factory, a smart city, an autonomous driving vehicle—a lot of data is being generated there and collected there at the location of the sensor.”
Willis and Hurzook’s concentration on the edge rather than the centre to find solutions for processing previously inaccessible data overcomes the problem of clogged networks that can result from centralized approaches. By relaying too many sensors’ data, as it were, a centralized topology can lead to all of the sensors trying to contact the same remote server at the same time. However, not every sensor that is IoT-connected needs access to a remote server.
Instead, with progress in applying control theory in places like sensor fusion engines or inertial navigation systems, engineers can relieve some of that burden off of the cloud. Since IoT needs quicker processing to make quicker decisions, localized processing at the edge is somewhere that industry ought to watch—especially now at the dawn of 5G.
Additionally, the influences of IoT on developers and manufacturers of chips happens in two phases: the first being how their products or processes can better serve IoT in general and the second being how IoT can better serve those same products or processes.
Willis expanded: “What is unique at Arrow, in our Five Years Out strategy, is that our Sensor to Sunset mentality allows us to work with companies to start at the design phase—at the chip level—all the way through product lifecycle, and, finally, obsolescence.”
The mentality Sensor to Sunset harks a similarity to digital twins. Picking up steam, particularly over the last five years, digital twins are the virtual copies of physical assets that allow for remote observation and cataloguing of sensor data that describe the states of a physical asset throughout its lifecycle.
The act of digitizing a copy of a physical asset through harvested data subdues the need to spend valuable resources on prototyping. Models and scenarios can be virtually tested on the twin to mimic real-world stresses. It’s easy to see where digital twins can speed up time-to-market. Machine learning even allows for the training of smarter digital twins in places such as the development autonomous vehicles .
Opportunities in Privacy, Security and AI
At the Driving Dreams workshop, speculation and excitement surrounding last month’s CES 2019 in Las Vegas lingered in the air, and it wasn’t just hype. Serious discussions about the social weight to decisions in the development of products—such as those with respect to IoT sensors and machine learning—were shared between attendees, panelists, and organizers alike.
“Consumers are giving up some of their freedoms for the convenience of IoT,” said Hurzook. “What they often don’t realize is that some of these cloud-connected systems are going back to central server, making the decisions and affecting things in your home. They think they are getting the convenience, but they don’t realize all the touch points of where your data is actually winding up.”
But concerns about privacy, security and data collection—as Barter explains—don’t end at consumer products.
“It’s one thing to worry about the privacy of our personal devices, but it’s another thing to worry about the privacy and security of our nuclear power stations, driverless cars, or heart-rate monitors,” Hurzook said. “I think there are very few things that are more important today than privacy and security, and it’s a key topic of conversation.”
It was clear that these discussions were not intended to scare consumers into feeling they needed to purchase security solutions, but rather to be knowledgeable about the risks and to use what solutions developers and manufacturers often already provide. Sentiment between developers and trend followers reiterate the necessity of good online security and hygiene among everyday consumers. With IoT bringing fridges and thermostats peripherally online, consumers deserve to have the information to and ability to protect their privacy as best they can.
According to the panel, it isn’t unusual for security features that are provided by product developers and manufacturers to go ignored by the average consumer. Sometimes, demand among consumers is lacking in the protection of their own data. Regardless, industry is keeping a close eye on what consumers of IoT products want. If consumers want more security and privacy, then the best way to make that sentiment known to developers is by putting their money where their mouths are.
At the same time, the panel agreed, the gap between privacy and commercial exchange of data serves as an extra opportunity for engineers and security specialists to innovate new products .
Additional opportunities for innovation mentioned during the panel can be found in better machine learning so that it doesn’t develop biases due to biased data sets. Fesiak mentioned oversight and better data hygiene as ways to ensure that new datasets don’t corrupt the next generation of AI and compound any problems.
Barter offered an example to help illustrate Fesiak’s point about bias in AI. “ Orchestras have changed in the last 50 years,” he said. “They were historically male. A couple of decades ago, for the first time, new potential hires were placed behind a physical barrier during their auditions. All of a sudden, we saw women begin to make up 45 percent of the new hires for such orchestras.”
Talent scouts and casting managers were able to listen to skill instead of letting their biases influence them.
“We need to come up with analogies to those physical barriers in artificial intelligence systems to help get biases out of the equation,” Barter concluded.
Finding these analogies will be an interesting next step for research and innovation in machine learning as well as AI’s relationships with IoT, autonomous vehicles, and smart spaces.
What’s Next?
So, what’s next for the 2019 Driving Dreams Expert Series? The second of the six workshops is scheduled for Mar. 21 and focuses on the theme of commercialization with topics such as “go-to market strategies, financing options, global supply chains, B2B & B2C,” according to the Driving Dreams website . The second workshop will also take place at Arrow Electronics’ office in Mississauga, Ont.