How AI and robotics redefined project delivery in 2024

A man with glasses smiles while using his laptop, reflecting a positive and focused work environment.

The pace of innovation in artificial intelligence (AI) showed no signs of slowing in 2024. From critical frameworks addressing the risks of rogue AI to humanoid robots redefining the future of work, the past couple of months have been a whirlwind of breakthroughs and thought-provoking developments. Our focus should always be on how these technologies intersect with project delivery, bringing new possibilities – and challenges – into play.

Here’s my take on the most compelling stories from the past few months, taken from the Project Flux weekly newsletter.

Understanding the threat of rogue AI

Let’s start with one of the most urgent discussions: the growing need to manage the risks associated with rogue AI. Anthropic’s Dario Amodei has introduced a framework called AI Safety Levels (ASL) to address this. It categorises AI systems from ASL-1 (low-risk) to ASL-5 (existential threats) and adopts a smart “if-then” approach. This means that as AI systems become more capable, stricter safety measures are automatically triggered.

Right now, we’re sitting at ASL-2, where AI systems can’t autonomously cause harm. But ASL-3 is just around the corner. That’s where things start to get real. AI could assist malicious actors or even pose risks through unintentional misuse. Beyond that, ASL-4 and ASL-5 introduce the possibility of autonomous AI systems acting beyond human control.

Why does this matter for project delivery? Well, as AI becomes more integrated into project management, the risks of misuse or error grow exponentially. Imagine an advanced AI system meant to optimise structural engineering being exploited to reveal vulnerabilities instead. The ASL framework offers a blueprint for balancing innovation with precaution, ensuring we continue to leverage AI responsibly.

Beware the incoming wave of AI agents in 2025

AI agents – autonomous systems powered by large language models – are gearing up to take centre stage in 2025. Think of them as ChatGPT’s more ambitious cousins. Unlike standard AI tools, these agents are proactive. They are capable of making decisions, integrating with tools and executing tasks in real time. This technology was recently introduced into Microsoft Copilot. Take the opportunity to explore and experiment with its features.

LangChain’s fascinating State of AI Agents report highlights just how quickly these systems are being adopted. Many organisations are already using AI agents in some capacity, with applications ranging from personal productivity to customer service. But in 2025, AI agents will move into high-stakes, hardware-integrated roles. Picture drones working in tandem with AI agents to manage multimodal networks or oversee construction sites in real time. For project delivery, this could mean:

dynamic resourcing: AI analysing project data and reallocating resources on the fly.
safety monitoring: sensors powered by AI agents identifying risks before they become incidents.
workflow automation: systems handling repetitive tasks, freeing teams to focus on strategic goals.

Of course, the challenge lies in rolling them out effectively. Businesses will need to address issues like bias, lack of transparency and operational oversight. But for those ready to adapt, the potential is transformative.

Dewan unveils AI-powered BIM breakthroughs

Autodesk University 2024 was buzzing with innovations. Dewan Architects + Engineers revealed its AI-enhanced building information modelling (BIM) tools. The standout development, Quible, integrates deep learning with BIM workflows, allowing real-time, data-driven design decisions. It seamlessly incorporates various data sources into BIM models, enhancing project visualisation and decision-making. Quible automates tasks like clash detection, 4D construction simulation and energy performance analysis, saving significant time and effort.

Quible also boosts collaboration among project stakeholders through a unified platform for sharing and updating project information. On the AMAALA Yacht Club project, Dewan’s AI-powered BIM and digital twins improved efficiency by 40%, redefining design and engineering collaboration.

For project delivery professionals, tools like Quible represent a shift towards more intelligent and collaborative workflows, with AI’s potential extending to on-site applications as well.

BMW’s humanoid worker

Robots in the workplace aren’t exactly a new idea, but BMW’s Figure 02 humanoid robot takes things to a new level. This isn’t your typical factory arm swinging bolts into place. Figure 02 can lift 400lb, move with human-like agility and adapt to its environment, thanks to advanced AI.

It’s remarkable how BMW has already implemented this technology in some of its factories, envisioning the robot not as a replacement for human workers, but as a collaborative partner. By handling repetitive or dangerous tasks, it allows humans to focus on creative and strategic work. In industries like construction and energy, the applications include:

lifting heavy materials on construction sites without risking injury.
conducting inspections in hazardous or unstable environments.
handling routine maintenance on infrastructure like bridges or oil rigs.

The bigger picture here is collaboration. As AI-powered robotics become more advanced, the lines between human and machine teams will blur, creating safer, more efficient workplaces. Future advancements will see these AI robots leveraging other AI tools, both in hardware and software, to further enhance their capabilities and adaptability.

Wrapping up

The past few months have been a powerhouse for innovation in AI and robotics. It’s clear that the future of project delivery is already here.

The challenge is to stay ahead of the curve. Whether it’s implementing safety frameworks like ASL, exploring the potential of AI agents or rethinking workflows with tools like Quible and robotics, the time to act is now.

As we move into 2025, the Project Data Analytics Task Force will continue to raise awareness through its meet-ups, events, newsletter and podcast. Project Flux will continue to explore these developments, helping you navigate the evolving landscape of AI and robotics.