Steve Rogerson walked the halls at Embedded World to find how the industry had recovered from the dark days of the pandemic.

The electronics industry has brushed itself down from the supply-chain and other problems that came from the pandemic and is now ready to face the challenges ahead, and the big one is artificial intelligence – not the hype but the reality.

That was one of the key messages that came from this month’s Embedded World show in Nuremberg, where it was nice to struggle walking the halls as the crowds returned to what seemed like pre-Covid levels. Most of those that had booked stands seemed happy as busy day followed busy day.

“Every company went through their supply chain challenges during the pandemic,” said Deepali Trehan, vice president at Intel’s Altera (www.altera.com) subsidiary. “We have come a long way and learned hard lessons so we never have to go through that again.”

She went onto say during the show’s c-level panel that though AI was complex the opportunities it provided kept on growing and it was important to stay ready for a changing landscape.

“Nobody can predict what will happen in the next five years,” she said. “You want the expertise of a big company but you need to be nimble and respond urgently.”

Thomas Böhm, senior vice president at Infineon (www.infineon.com), said the importance of software had increased dramatically over the past few years.

“Software is key for innovation,” he said. “And innovation from the outside is important, so we work with hardware and software partners. You have to be ready to destroy yourself partially, unlearn some habits and create new innovations.”

Kevin Xu, director at Chinese firm Biwin Storage Technology (www.biwintech.com), said AI had become the most significant topic, and that would continue.

“To keep up with AI, you have to invest in a lot of new technologies,” he said.

Deepali agreed, saying that three quarters of customers want to add AI to their designs in the next four to five years.

“You need to keep that journey as simple as possible for them,” she said. “You have to focus on ease of use for the developer experience. We are still in the early innings for AI. Our customers intend to add AI but only a few know exactly what they are going to do.”

Later in the week, I talked with Florian Denzin, product strategist at Telit Cinterion (www.telit.com), about this and he said AI at the edge would change IoT and the goal was to combine AI and connectivity in the best way.

“We are trying to show with reference designs how these can work together,” he said.

Edge AI, he said, was today focused on vision applications. Vision applications existed already but needed the power efficiency to bring them to the edge.

Among the main AI announcements at the show, two were powered by Qualcomm (www.qualcomm.com). The first came from Fibocom (www.fibocom.com), which unveiled some Linux-based edge AI products using Qualcomm technology (www.iotm2mcouncil.org/iot-library/news/iot-newsdesk/qualcomm-powers-fibocom-linux-based-smart-modules).

The second was from Advantech (www.advantech.com). The Taiwanese company announced a partnership with Qualcomm (www.iotm2mcouncil.org/iot-library/news/iot-newsdesk/advantech-and-qualcomm-to-bring-ai-to-edge) that involved edge AI, paving the way for AIoT applications.

“Edge AI is important to every industry,” said Miller Chang, president of embedded IoT at Advantech. “It will influence many, many emerging markets.”

AI has been having some mixed press lately, with some fearing it is becoming too powerful while others delight in how the recent advances can make everything so much better. Those on the negative side have scored some wins with legislation in the USA and Europe implementing restrictions on the use and development of AI.

Californian software company Iterate.ai gave one of the more measured responses. Its vice president Magnus Tagstrom said: “As the USA, UK and EU establish key guidelines for AI regulation, we find ourselves at the beginning of an important shift in technology governance. The memorandum signed between the USA and the UK, together with the EU’s significant legislative strides, highlights a collective dedication to responsibly unlocking AI’s potential. It is crucial that we continue to enhance international cooperation and dialogue to ensure AI’s development serves the greater good of humanity.”

He said the challenge ahead would be to update rules and regulations as AI continues to evolve.

“The truth is, as we venture into these unknown territories, the essence of our humanity and ethical principles are key,” he said. “It’s vital that we amplify international cooperation and dialogue to ensure AI’s trajectory aligns with human dignity and liberty. Facing this challenge head-on, as a global alliance, is the only way to sculpt an AI landscape that is secure, just and deeply rooted in respecting human rights.”

The timing of the advances in AI technology really could not be better. We all had a few rough years when Covid hit and needed something to give our industries a boost. AI, and its deployment at the edge of IoT networks, is looking like just what we wanted, when we wanted it.