Click a track to filter sessions
Select a day
- Tuesday
- Wednesday
- Thursday
-
11:00am
Edward Bennett Unreal11:00am - 11:45amRoom 4Future-Proofing Game Dev: Upcoming Trends & Strategies
Tuesday 11th July11:00am - 11:45amRoom 4This talk focuses on future trends in gaming, ensuring your career and game are prepared for what's ahead.
The talk will cover:
- Shift towards procedural generation (PCG) in game development.
- Increasing interconnectivity of IPs, with game assets used in virtual production and more.
- Exploding demand for content and supporting tools.
- Verse and why it will be important in a metaverse age and why they should consider learning it.
- Advantages of publishing to Unreal Engine's Fortnite (UEFN) for visibility and leveraging platform work.Session Takeaway
12:00pm
Séamus Ó Buadhacháin Failbetter Games12:00pm - 12:45pmRoom 3Testing Branching Narrative at Scale: Lessons from Mask of the Rose
Tuesday 11th July12:00pm - 12:45pmRoom 3Projects with complex narratives can be unduly time-consuming to test, requiring multiple human playthroughs for each story choice. Mask of the Rose, a game with thousands of individual choices and millions of unique narrative routes, presented developers Failbetter Games with a serious testing challenge. This talk covers how we addressed the challenge of testing and debugging a branching narrative at tremendous scale, with a guidable random playthrough system that evolved from finding showstopping quest bugs to evaluating the flow of procedurally-generated conversations.
Session Takeaway
- The challenges of testing big narrative games
- The benefits and downsides of manual testing vs automated testing
- Challenges and pitfalls in developing automated testing for narrative-heavy games
5:00pm
Vicky Bloom Astrid Entertainment5:00pm - 5:45pmRoom 2Revolutionising Game Worlds: Development in Emergent AI
Tuesday 11th July5:00pm - 5:45pmRoom 2Dive into the innovative world of Utility AI and its transformative impact on non-player character (NPC) behaviour in video games with Dr. Vicky Bloom. She brings her expertise in AI and game development to introduce the fundamental concepts of Utility AI. A system that empowers NPCs to make lifelike decisions, providing a significant advancement over traditional AI by providing more dynamic and emergent interactions.
Vicky will introduce Utility AI's basics, including its decision-making models and scoring systems, which enable NPCs to exhibit varied and complex behaviours. She'll share insights from her experience as Lead Software Engineer at Astrid Entertainment, using practical examples from their current game to demonstrate how Utility AI can transform NPC interactions and game worlds.
The session will cover the integration of Utility AI into Unreal game development, focusing on rapid development with plugins and best practices. Attendees will learn how to start with Utility AI, from concept to execution. They will gain a foundation for applying this technology in their projects, paving the way for the next generation of interactive entertainment.
Whether you are a student eager to learn about the cutting edge of game AI, an entry-level developer looking to enhance your skill set, or simply curious about the future of game design, this talk will provide you with a comprehensive introduction to one of the most exciting advancements in game development technology. Don't miss this chance to see how Utility AI can animate your game worlds in new, exciting ways.
Session Takeaway
- An understanding of Utility AI basics and how it creates dynamic NPC behaviour with real-world examples.
- Learn how to seamlessly incorporate Utility AI into Unreal projects, emphasising plugin integration and best practices.
- Gain practical insights from applying Utility AI in Astrid Entertainment's current project, inspiring the creation of captivating game worlds.
-
11:00am
Christian-Peter Heimbach Actioncy Tech11:00am - 11:45amRoom 4Practical Machine Learning for Game Developers, Publishers and Studios
Wednesday 12th July11:00am - 11:45amRoom 4A Journey from Theory to App.
Demystify machine learning’s complexities and showcasing its practical applications in game publishing. Starting with a foundational overview of ML concepts, Heimbach shares his journey of fine-tuning a machine learning model, offering a candid post-mortem analysis of the challenges faced and lessons learned. Attendees will gain insights into the iterative process of ML model development, from data gathering and model training to evaluation and fine-tuning, with a focus on actionable strategies and use cases. This talk aims to equip game developers with the knowledge and tools to leverage machine learning in their own projects.
Session Takeaway
- Foundational Understanding: Grasp core ML concepts and how they can be applied to game development.
- Practical Insights: Learn through a real-life case study of developing a machine learning model, including selecting the right algorithms, dealing with data, and overcoming obstacles.
- Actionable Strategies: Discover effective techniques for model evaluation and fine-tuning to enhance game features, such as AI behaviour, content generation, and player experience personalization.
2:00pm
Lisa Ridley Ubisoft Reflections2:00pm - 2:45pmRoom 3Assassin’s Creed Nexus VR: Delivering A Performant Living World in Mobile VR
Wednesday 12th July2:00pm - 2:45pmRoom 3This session will explore the ways that the team behind Assassin’s Creed Nexus VR worked to bring a AAA experience from a much-loved franchise into mobile VR, and used the limitations of the Quest 2 as an opportunity to think innovatively to build an immersive living world.
It will include tangible examples with gameplay footage and performance numbers, and will also explore how the team looked into the franchise’s history (Assassin’s Creed Unity) to elevate the experience.
Session Takeaway
- Understand the challenges and trade-offs involved in achieving AAA quality in mobile VR
- Learn techniques to boost performance without sacrificing NPC fidelity
- Discover what it takes to achieve realistic NPC reactions in VR
4:00pm
Ihor Pavlovskyi VirtuosPeter Sikachev VirtuosOstap Trush Virtuos4:00pm - 4:45pmRoom 2Directional Light Shadows Compute Time-Slicing Optimization
Wednesday 12th July4:00pm - 4:45pmRoom 2In UE5, sun directional shadow calculations are expensive, even when using predominantly Nanite assets. Having a day/night cycle and recomputing the sunlight directional shadow every frame is not affordable for most titles.
In our presentation, we outline the problems that Virtual Shadow Maps in UE5 have with the day/night cycle and propose one solution that consists of distributing shadow computation generated by sunlight between multiple frames.
We present potential performance gains that reach up to 70% on shadow computation as well as talk about visual losses that you may get with this algorithm.We believe that by sharing this idea with developers, we can provide a better understanding of how shadow mapping can be optimized in UE5, as well as in other engines. Furthermore, our goal is to encourage developers to optimize their graphic bottlenecks, so they can gain more frame time budget for other assets and technologies in their games.
Session Takeaway
- Low-level understanding of Virtual Shadow Map (VSM) Clipmap technology in Unreal Engine 5 (UE5) and its limitations
- How to use time-slicing to optimize VSM for a dynamic day-night cycle game in UE5
5:00pm
Louis Larsson-De Wet The Chinese Room / Sumo Digital5:00pm - 5:45pmRoom 3Still Wakes the Deep: Tentacles and Water
Wednesday 12th July5:00pm - 5:45pmRoom 3Still Wakes The Deep is set on an oil rig in the North sea off the coast of Scotland circa December 1975. It’s “The Thing” on an oil rig. The sea, and water in general, plays a big role in our game. Not only are there sprawling vistas of the sea, but you have standing pools of water, flooding rooms and claustrophobic swimming sections.
Water is not your only enemy on the rig – something else came onboard.
For the first part, I will talk about our water implementation, which includes real-time water simulations.
And then in the second part, I will talk about our monsters – in particular procedural animation and how to make them move in a realistic way with their tentacles through a complex environment.
Session Takeaway
- Understand the challenges of having water in a game
- How to implement different kinds of water, including real-time simulated water and dirty water
- See how to make a monster move realistically through a complex environment.
- Niagara Systems and Control Rig in Unreal Engine
-
12:15pm
Antonia Forster Unity12:15pm - 1:00pmRoom 1Getting Started Developing for Apple Vision Pro Using Unity
Thursday 13th July12:15pm - 1:00pmRoom 1In this talk, you'll learn the fundamentals of building to the Apple Vision Pro (or the Apple Vision Pro Simulator) using Unity. We'll explore the different modes of applications you can build for the device and learn about Unity's visionOS support (including Unity's PolySpatial package for visionOS development). You'll learn about the tools and workflows required to deploy to the device, and the samples and templates available to help you get started. This session is your gateway to unlocking the potential of creating immersive applications for Apple's innovative platform.
Session Takeaway
- The different types of applications you can build forApple Vision Pro
- Requirements for visionOS development
- The different types of applications you can build for Apple Vision Pro
4:00pm
Dominik Jeurissen Queen Mary University of London4:00pm - 4:45pmRoom 3LLM Agents for QA - Potential & Limitations
Thursday 13th July4:00pm - 4:45pmRoom 3With tight deadlines and a constantly evolving game, properly testing a game is challenging. Using AI agents to simplify this work sounds promising, but machine learning is often too slow, and manually implementing the agents takes time. As such, one particularly exciting application for QA is to use Large Language Models (LLMs) as zero-shot game-playing agents. LLM-based agents can play games without pre-training, making them a valuable asset to test a constantly changing game. But how well do they play games? What are their strengths, and what do they struggle with?
In this session, we will review how to implement zero-shot agents with LLMs and show examples of existing LLM-based game-playing agents. We will also show that although these agents have many limitations, they have the potential to be a valuable tool for QA to automate many repetitive tasks.
Session Takeaway
- An overview of the cutting-edge research on LLM-based zero-shot game-playing agents.
- Learn what these agents can do well and what their limitations are.
- Practical tips on how to utilize LLM agents as QA tools.
