Top 7 Robotics Companies (2025): Revolutionizing Automation
How AI In IoT Is Powering Sustainability
Nikhil Jain is a Senior Partner Technology Manager at Samsung SmartThings Inc, with over 12 years of experience in the smart home industry.
gettyAmid a world striving for sustainability, the convergence of AI and IoT is becoming a key enabler, helping industries optimize resources, cut emissions and build intelligent systems.
The AI-IoT SynergyIoT devices serve as digital senses, gathering data from homes, factories, farms and cities. AI acts as the brain, analyzing this data to generate insights, predict trends and make real-time adjustments.
For example, Hong Kong's CLP Power used AutoGrid's platform to invite 950,000 smart-meter users to cut usage during peak hours. About 70% responded, reducing electricity use by 410,000 kWh in just four hours, avoiding roughly 160 tons of CO₂ emissions. This shows how demand forecasting and automation can ease grid load and reduce emissions.
Taking a further look at how this works:
• Intelligent Energy Optimization: AI models can forecast peak energy demand and automatically adjust the consumption of non-essential assets, enabling demand response strategies that reduce reliance on carbon-intensive energy sources.
• Predictive Maintenance: AI identifies early signs of equipment failure from industrial machines to smart thermostats allowing issues to be addressed before downtime occurs, reducing both energy waste and productivity losses.
• Autonomous Optimization: AI continuously fine-tunes lighting, HVAC, and irrigation systems to conserve both energy and water.
Integrating AI and IoT in manufacturing alone reduced energy use by up to 20% and cut emissions by 6.1% across North America.
How To Measure Sustainability: From Data To ActionTo fully realize the environmental benefits of AI and IoT, organizations must track the impact of their deployments. Key metrics include:
• Energy Consumption: Power usage of both devices and AI models.
• Carbon Footprint: Emissions associated with model training, deployment and ongoing operations.
• Efficiency Gains: Improvements such as crop yield per unit of water or energy saved per task.
• Environmental Parameters: Indicators like indoor air quality, water usage and waste levels.
Emerging tools such as Microsoft's Sustainability Calculator and the ML CO2 Impact Calculator help companies quantify and report these metrics, ensuring that innovation goes hand in hand with improved ESG performance.
Sector Impact Snapshot 1. Smart ManufacturingAI-IoT integration enables factories to move from reactive to predictive systems. Machines are self-monitoring, allowing maintenance only when necessary. Real-time inventory tracking supports circular economy models by minimizing overproduction and obsolescence through accurate tracking of returned units and their conditions, helping lead to leaner operations and lower emissions.
2. AgriculturePrecision farming, 3. Energy And Smart Grids
AI-IoT solutions facilitate renewable energy integration, real-time load balancing and reduced transmission losses. Cities increasingly rely on AI to maintain electric grid stability while expanding the use of distributed energy resources (DERs), such as rooftop solar panels.
4. Smart Cities And Supply ChainsFrom intelligent traffic systems to adaptive street lighting, AI-IoT implementations are cutting urban energy consumption by up to 30%. In supply chains, these technologies optimize logistics, reduce waste and enhance transparency, leading to more sustainable global operations.
Barriers And ChallengesDespite its potential, the adoption of AI-IoT solutions is slowed by several key hurdles:
• Security And Privacy: IoT devices often have vulnerabilities, and AI models rely on sensitive data, raising concerns about data breaches and misuse.
• Interoperability: A fragmented ecosystem of devices and platforms makes seamless integration difficult.
• Connectivity: Reliable, high-speed networks are essential for real-time data and control, yet many regions lack adequate infrastructure.
• Skills Gap: There is a shortage of professionals skilled in both AI and IoT, limiting deployment and innovation.
• Explainable AI: Many AI models operate as "black boxes," which is problematic in critical sectors like healthcare and energy where transparency is vital.
To Make A Difference, Break Attitude Barriers To AdoptionTo truly unlock the potential of AI-IoT solutions, organizations must confront key adoption challenges head-on. Targeted reskilling programs, partnerships with academic institutions and interdisciplinary training that combines data science, embedded systems and sustainability are critical to closing the AI-IoT talent gap.
At the same time, ensuring security and privacy requires a robust, zero-trust architecture supported by end-to-end encryption and proactive governance that builds trust across every layer of the technology stack.
Finally, while these technologies deliver quick wins in efficiency and emissions reduction, organizations must also account for their full lifecycle impact. This includes evaluating the energy required to train AI models and manufacture IoT hardware to ensure that long-term sustainability gains truly outweigh the environmental costs of conventional approaches.
Sustainability Within AIIronically, AI itself can be energy-intensive. Training large models consumes significant electricity and even water, making it essential to keep net-positive outcomes in focus. To address this, organizations are increasingly adopting strategies such as:
• Edge AI: Processing data closer to where it is generated reduces transmission costs and energy usage.
• Green Model Design: Developing smaller, more efficient models that require less computation and shorter training cycles.
• Lifecycle Assessments: Evaluating both the environmental benefits and trade-offs qualitatively and quantitatively associated with deploying AI-IoT systems.
The Future: Smart Sustainability On The HorizonThe path forward will be shaped by several key enablers:
• Standardized metrics for environmental impact reporting
• Blockchain integration to enhance transparency and traceability in sustainability efforts
• Regulatory frameworks that incentivize tech-driven environmental solutions
• Multi-sector partnerships to accelerate real-world implementation
• Next-generation infrastructure supporting edge computing and 5G-enabled IoT
This is a pivotal moment for technologists, startups and global enterprises. Those who strategically adopt AI and IoT—not just for innovation, but for measurable impact—can lead the charge toward a more connected and sustainable future.
ConclusionAI in IoT is not just revolutionizing how we live and work; it's also transforming how we care for our planet. By combining intelligence and connectivity, we can create systems that are both smart and sustainable.
As adoption grows, the opportunity extends beyond improving products—it's about building a better, more sustainable world.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Multimodal AI For IoT Devices Requires A New Class Of MCU
The rise of AI-driven IoT devices is pushing the limits of today's microcontroller unit (MCU) landscape. While AI-powered perception applications—such as voice, facial recognition, object detection, and gesture control—are becoming essential in everything from smart home devices to industrial automation, the hardware available to support them is not keeping pace. The challenge? The broad 32-bit traditional MCU install base cannot handle the demands imposed on them by AI-ready workloads.
While new solutions (such as AI MCUs) are being launched by semiconductor vendors, the overall experience is still less than ideal. Architectures are too rigid, software and tooling continues to be proprietary, and solutions are overly complex.
In addition, many AI-enabled devices on the market today are repurposed from silicon originally designed for other applications—mobile, cloud, automotive or general-purpose embedded computing. Such architectures, while powerful, are not optimized for the ultra-low-power, always-on operation required by IoT devices. The result is a fragmented AI ecosystem, where designers must choose between low, medium, or high AI processing capabilities—often leading to trade-offs in performance, power efficiency, and cost.
Compounding the issue is the inefficiency of existing system architectures. Many AI MCUs rely on rigid, inflexible designs that are unable to balance compute power with energy efficiency. In an IoT environment where devices must operate for months on a single battery charge, this mismatch leads to unnecessary power consumption and limits the potential for AI at the edge.
To overcome these challenges, the industry needs a new class of MCUs that blend intelligent sensing with AI-accelerated compute. These next-generation MCUs must deliver high-efficiency AI processing, scalable performance, and energy optimization tailored for always-on, low-power applications.
Fig. 1: Multimodal context-aware compute.
To address this, Synaptics introduced the Synaptics Astra SR-Series platform of context-aware AI MCUs for IoT devices. The Astra platform integrates hardware, open software, dev kits and ecosystem partnerships.
Introducing a multi-gear architecture for intelligent IoT AI processingThe first-generation SR100 Series of high-performance MCUs introduce a novel three-tiered architecture designed to optimize AI processing for IoT devices. Unlike traditional MCUs that either remain fully powered or completely idle, the SR110 dynamically adjusts its compute power based on real-time system demands. This context-aware computing enables ultra-low-power (ULP) operation while maintaining high-performance AI capabilities when needed.
At the core of this design are three distinct compute domains, or "gears," that operate at different power levels to balance energy efficiency and AI processing performance.
ULP Always-On Domain: Continuous activity monitoring at ultra-low power
The always-on (AON) domain is responsible for continuously monitoring the environment for variations, even when the primary CPUs are in sleep mode. This ensures the system remains responsive without draining battery life.
This domain is designed to detect both vision- and audio-based events, such as motion, sound patterns, or changes in lighting. It can generate wake-up triggers based on pre-programmed detection parameters.
Another feature of the SR100 series is low power pre-roll. As events are monitored, they are stored on device. And when a trigger happens, the trigger event and the events leading to it can be sent for deeper analysis to the next domain.
Efficiency Domain: Low-power AI processing for real-time event detection
The efficiency domain is responsible for handling initial AI inferencing after an event is detected. It consists of an Arm Cortex-M4 MCU running at 100 MHz and a custom micro-NPU (Neural Processing Unit) from Synaptics, delivering up to 10 GOPS of AI inferencing power.
When a wake-up event is triggered (such as a detected object or audio cue), the compute elements in the efficiency domain process the data with lightweight AI models to determine the nature of the event. This enables real-time object detection, sound event detection, and other basic AI tasks, while maintaining low power consumption.
If additional processing is required—such as higher-resolution facial recognition or complex AI inferencing—the system escalates to the performance domain.
Performance Domain: High-compute AI acceleration for advanced processing
For more demanding AI tasks, the performance domain is activated. This domain provides significantly higher processing power, making it suitable for computationally intensive applications such as facial recognition, body pose estimation and advanced object detection and classification.
The performance domain consists of an Arm Cortex-M55 MCU, with Arm Helium extensions running at 400 MHz, providing high-speed AI execution within an MCU framework and a high-performance NPU (Arm Ethos U-55) also operating at 400 MHz, delivering up to 100 GOPS of AI inferencing capability.
This novel, tiered processing structure ensures that only the necessary compute power is used at any given time, dramatically improving energy efficiency without compromising on performance.
The future of context-aware AI computingThe SR100 series' intelligent gearing algorithms dynamically shift between these compute domains based on the system's needs. This context-aware AI computing improves scalability, energy efficiency and standardized development. This platform approach is more flexible, and leads to more-standardized development practices, essential to an IoT AI space that's evolving rapidly.
With a rich set of I/O and peripherals, including MIPI-CSI, lightweight ISP, USB, serial interfaces and security features, the SR100 MCU delivers a versatile, power-efficient, and highly programmable solution for multimodal, AI-enabled IoT devices.
Qualcomm Opens New AI R&D Center In Vietnam
In sum – what to know:
AI R&D hub – The U.S. Chipmaker opened a new AI research center in Vietnam focused on generative and agentic AI for mobile, XR, automotive and IoT use cases.
Supports Vietnam's tech ambitions – The center aligns with national goals in AI, semiconductors and digital transformation, helping build local talent and advance the country's position in regional innovation.
Strengthens U.S.–Vietnam ties – The move deepens strategic tech cooperation between the two countries.
U.S.-based tech giant Qualcomm officially opened a new artificial intelligence research and development (AI R&D) center in Vietnam, according to local press reports.
The new Qualcomm facility, staffed by scientists and AI experts in both Hanoi and Ho Chí Minh City, is part of company's global AI research division. The center will focus on advanced generative and agentic AI, with applications in smartphones, PCs, XR, automotive and IoT, according to the report.
At the launch event in Hanoi, Vietnam's deputy minister of Science and Technology, Le Xuan Dỉnh, noted the new AI facility would boost nation's AI research capabilities, help train up skillsets of local AI experts and contribute to Vietnam's economic development.
Meanwhile, Thieu Phuong Nam, country director of Qualcomm Vietnam, Cambodia and Laos, said that Qualcomm's activities in the Asian nation are in line with Vietnam's national strategies on AI, semiconductors and digital transformation.
Qualcomm currently operates two offices in Hanoi and Ho Chí Minh City and launched its first Southeast Asian R&D center in Hanoi in 2020. It also runs the Qualcomm Vietnam Innovation Challenge, a program supporting local startups through funding, technical guidance and IP assistance.
The report also noted that Vietnam has set an ambitious goal to become one of Southeast Asia's top three nations for AI research and development by 2030.
In April, Qualcomm had announced the acquisition of Hanoi-based generative AI research specialist MovianAI, a former division of VinAI, and a part of Vietnamese industrial tech conglomerate Vingroup.
The deal for MovianAI will see the division's founder, Dr Hung Bui — formerly of Google DeepMind — join Qualcomm, and continue to lead the team in Hanoi, which has been described by the California chipmaker in a press statement as a "powerhouse" generative AI team pushing "boundaries" with "customized AI models and engineering." Beyond generative AI, its research focuses on machine learning, computer vision, and natural language processing. Qualcomm said the combination will "expand its ability to drive extraordinary inventions."
"We are ready to contribute to Qualcomm's mission of making breakthroughs in fundamental AI research and scale them across industries," said Bui.
Qualcomm has worked closely with the Vietnamese technology ecosystem for 20 years, it said. The company continued: "Qualcomm's innovations in the areas of 5G, AI, IoT and automotive have helped to fuel the extraordinary growth and success of Vietnam's information and communication technology industry and assisted the entry of Vietnamese companies into the global marketplace."
Qualcomm Technologies and Emirati-based telco e& UAE have recently announced a strategic partnership aimed at advancing the development and commercialization of next-generation connectivity, 5G and edge AI technologies.
Under the partnership, the chipmaker and e& will work together on:
-5G edge AI gateways for industrial and enterprise use, helping major sectors implement AI and other technologies closer to the data source to boost efficiency, performance and connectivity.
-Edge AI devices, including PCs and extended reality (XR) gadgets
Additionally, Qualcomm Technologies said it plans to leverage its new Qualcomm Engineering Center in Abu Dhabi to support this initiative. The center will help explore and test new applications to expand the use of 5G and edge AI in sectors such as energy, manufacturing, logistics, retail and transportation.
Juan Pedro Tomás Juan Pedro covers Global Carriers and Global Enterprise IoT. Prior to RCR, Juan Pedro worked for Business News Americas, covering telecoms and IT news in the Latin American markets. He also worked for Telecompaper as their Regional Editor for Latin America and Asia/Pacific. Juan Pedro has also contributed to Latin Trade magazine as the publication's correspondent in Argentina and with political risk consultancy firm Exclusive Analysis, writing reports and providing political and economic information from certain Latin American markets. He has a degree in International Relations and a master in Journalism and is married with two kids.
Comments
Post a Comment