nitrosodiumfmpsequel

[nitrosodiumfmpsequel]

Planned Research into Specialist Practice

Next week, before I start actually developing Robot Shephard, I will do a load of research into games that are similar to my idea; i.e. puzzle games that focus on commanding NPCs. I'm going to do my proposal now, but these are future sources that I plan to draw from down the line.

Then I also have this game to look at, Terra Nil, which has an interesting solarpunk vibe. That will be good to research for the level aesthetics.

Finally, I will also watch some lectures from the GDC (Game Developers Conference), to learn more about relevant aspects of game design. I went to their channel and picked out two that seemed to be interesting; one about designing puzzles and another about color theory.

[nitrosodiumfmpsequel]

Choosing my Idea

I have considered the pros and cons of each concept - the actual basis of the game, and what the development process would probably entail.

METAL JUNGLE - Pros

I haven't had a proper look at procedural generation yet, so that could be something interesting to look at.

The procedural generation creates replay value, and would let me create a lot of content in a short time.

The top-down perspective would let me produce more 2D sprite work, which I enjoy.

METAL JUNGLE - Cons

The "facility overrun by mutants" is a fairly overdone trope.

If I want to include things like a complex inventory system, the code could get very complex, taking away dev time from polishing aspects of the game.

ROBOT SHEPHARD - Pros

The Lemmings-esque mechanic of guiding NPCs through a level, combined with the first-person aspect, is a unique concept and it seems like a good idea to expand on it.

I usually play first-person games, so I have a good idea of what makes them fun, compared to dungeon crawlers which I do not play so much.

The relatively light subject matter means that Robot Shephard will be a more universally accessible experience that anyone can play.

ROBOT SHEPHARD - Cons

I have already done many first-person projects (Sinister, Fulcrum, etc) so it could come across as derivative.

STEEL WANDERER - Pros

I haven't used the Top Down template before, so it could prove interesting.

STEEL WANDERER - Cons

I would have to make a fairly large map to convey the open world elements.

Conclusion

After a lot of consideration, I have decided to choose Robot Shephard as the idea to make for the FMP. This was something of a difficult decision, as I think all the ideas have their merits, but Robot Shephard seems the most innovative with its puzzle gameplay. Metal Jungle could definitely be fun, but I feel like the general concept has been done many times before. I also can't create a curated experience like I would want to with a static map. As for Steel Wanderer, I feel as if the gameplay would be somewhat lacking - just roaming through an empty map, collecting pieces of junk and then reading your character's thoughts on each one. Robot Shephard on the other hand seems like it will let me do a lot of interesting things with the concept - just based on the interaction systems between the player and the robots, I could make lots of different puzzles and situations to maneuver through. The solarpunk aesthetic could also be interesting to explore, since there are many themes of the style (renewability, plant life, et cetera) that could be focused on.

[nitrosodiumfmpsequel]

Steel Wanderer - Technical Feasibility

This idea has the most "uncharted territory" out of the three. I want to use the Top Down template for the game, which is one that I haven't used for a project before. I have had a little play around with it in my own time, but I'm not as familiar with it as I am with the First Person or Third Person templates.

I looked at this video, and found it very useful. It's for UE4, but that isn't too different from UE5 apart from how it looks visually. From what I can see, the template has a TopDownController blueprint and a TopDownPlayer blueprint. The Controller is where all the input and complex code is - it essentially registers when and where the mouse is clicked, and then instructs the Player to move there. So you don't actually have much control of your character in the typical sense. This could make more complex code difficult, i.e. where would you store important variables like health, the Controller or Player? You don't want to have to cast to one or the other every time an important event happens, since that'll just chew up processing power.

I'd probably be fine figuring this sort of thing out on my own, but it might mean that I don't get to spend as much time polishing the game, since I'd be learning new stuff. That's not to say the game wouldn't be functional, but it would be like the difference between Overdeath and Fulcrum for example. The former is me trying out a first person game for the very first time, and mostly just experimenting with what I can get from the genre conventions, meanwhile the latter is me fully immersed into the game style, letting me now push it as far as possible in terms of content.

[nitrosodiumfmpsequel]

Steel Wanderer - Ethical Research

Since Steel Wanderer doesn't feature any living creatures whatsoever, it should still qualify as PEGI 12 at the highest. And of course, no gambling, foul language, discrimination, et cetera. One thing that could be important to look at is how I handle the topic of war, since war is what wiped out humanity in the setting. I don't think the idea would glorify war at all - if anything, it's a fairly anti-war sentiment, seeing as nearly all life on Earth has been made extinct, leaving only AI constructs and a massive swath of flora. I suppose that, combined with the bombed-out ruins of city blocks which would comprise most of the map, could make it a tough game for people with any sort of war trauma.

Then, as for accessibility, this might be the most accessible of my games. Since I'll use the Unreal point and click template, it can be done with just one hand - left click for movement, and maybe right click for interaction. Then I'd of course steer away from all the inaccessible elements I've talked about before, relying solely on audio triggers for things, using colour-coding without an accompanying visual change like texture or shape, et cetera.

[nitrosodiumfmpsequel]

Steel Wanderer - Similar Concepts Part 3: Stray

Stray is an adventure game where you play as a cat in a post-apocalyptic world. You start off on the surface, before you fall into Walled City 99, a decrepit metropolis populated by robots called Companions. The Companions, as their name suggests, were built to be servants of humanity, but after society collapsed and humans became extinct, they continued living in the way that humans did. Throughout the game you can find Companions getting a haircut, even though they're made of metal and don't have hair, or eating at a restaurant despite not being able to digest food. This idea of artificial constructs surviving past the fall of humanity could be something to pursue in Steel Wanderer.

Obviously there is already the protagonist droid, but there could be other AIs and robots which have continued living. Maybe some sort of regulatory AI was connected to a wind farm, so it would be always online, and it sends off some sort of distress signal which is why the droid wakes up at the start of the game. That could be a long-term goal for the game, compared to the finding of artifacts, which could just be side content.

[nitrosodiumfmpsequel]

Steel Wanderer - Similar Concepts Part 2: The World Without Us

The World Without Us by Alan Weisman is a book that posits a thought experiment: what would happen to the world if humanity disappeared? It looks at various aspects of our society's impact on the world and sees how they would remain or be consumed by nature - cities, power plants, even long-lasting things like our production of plastic or petroleum. From what I've read online, Weisman predicts that the regular house would fall apart after 500 years, as water would leak into the roof and weaken the walls to the point of collapse. (This might be more of an America-specific prediction, however, since buildings are primarily wood over there, whilst over here they are made of the more endurant substance of brick.) Metal artifacts would probably be the longest lasting remnants of our existence; namely stainless steel, aluminium, and the various gold discs we have sent into space.

I could use this as a gameplay mechanic. Since the droid's goal is to figure out what happened to humanity, they could have to find artifacts and scan them for knowledge, which would give them hypotheses of humanity's nature. I could do some funny things with this. For example, the droid could find a frying pan and the analysis could be something like:

Item: STAINLESS STEEL DISC WITH HANDLE. Use: Unknown. Notes: Produces sonorous sounds when struck. Hypothesis: This device was used as a sort of call to prayer for the humans. Judging by its proximity to the sustenance storage appliances, it can be assumed that humanity worshipped food as a sort of deity.

Then again, that might take away from the bleaker atmosphere. Or it might add some much-needed levity, and turn the game into something of a dark comedy. It could be an interesting way to do things.

[nitrosodiumfmpsequel]

Steel Wanderer - Similar Concepts Part 1: Fallout

The original 1997 Fallout is an inspiration to me for many reasons. Firstly, it has the post-apocalyptic tone, but also the point-and-click control scheme that I want to try out for Steel Wanderer. There's also the general tone of the game, which is a lot bleaker than many of the other Fallout games. While Bethesda have made the Fallout series less serious, and more about the accelerated 1950s Americana aspects, the original game was almost grimdark in its presentation of the apocalypse. There's a bigger focus on the futility of war and the general hopelessness of the setting, shown in both the muddier colour palette and the strange industrial music, which can sometimes sound like warped nuclear sirens at times.

I could definitely lean into this with Steel Wanderer. Perhaps humanity wiped themselves out with a nuclear holocaust, which would give the game a distinct anti-war theme. I don't plan to include many animals either, so canonically it could have been some sort of bioweapon that killed off nearly all life, leaving just plants. This would make quite a depressing game, but it could be interesting to see what routes that would lead me down.

[nitrosodiumfmpsequel]

Steel Wanderer - Art Direction

Once again I'm going with the theme of overgrowth for Steel Wanderer, but this time focusing more on the abandoned elements. Canonically in this game, humanity has gone more or less extinct, and you as the droid will be waking up centuries after. The outlook is fairly bleak, and the only remnants of the human race are the shells of concrete and rebar we've left behind, now consumed by the sea of greenery.

While the broad subject matter is the same, it's a far emptier world than portrayed in my other ideas. Artificial fossils of bent steel pointing up from the blasted ground, remnants of houses now only filled with plant life and rust. It should provoke the player to ask questions as they navigate their drone through the map - "What happened to everyone? How long has it been since everyone disappeared?"

[nitrosodiumfmpsequel]

Game Idea 3 - Steel Wanderer

This idea focuses more on the 'abandoned' aspect of my research. It'd be a point-and-click isometric game, similar to Diablo or the early Fallout games, and you'd play as some sort of droid that has come awake after society's collapse. Everything is demolished and overgrown, and your goal is to explore around, manage your resources, and figure out where humanity has gone.

Here is a mind map to better explain my ideas:

[nitrosodiumfmpsequel]

Robot Shephard - Technical Feasibility

Compared to Metal Jungle, which would be made primarily using techniques I already know, Robot Shephard would be entirely new ground. Figuring out things like escorting NPCs, or directing groups of NPCs, will require looking at how other people have done it.

This video by the reliable Gorka Games shows the way I want to direct the robots around, via line traces. Some other mechanics could be interesting, like setting a point for the robots to recall to, or getting them to come to your side. This could probably be achieved with a "recall point" actor that gets spawned at the impact point of a line trace.

This is diagram of how I would probably do the line trace code. The code on the left is the player, and the code on the right is for the robot. I've done doors and elevators before in Sinister - they run off Timelines - so that shouldn't be too much of a hassle.

[nitrosodiumfmpsequel]

Robot Shephard - Art Direction

While Robot Shephard and Metal Jungle are using the same basic aesthetic of an overgrown facility, they look at different aspects of it. Metal Jungle is all the foul, dark elements of overgrowth; tangling vines, sharp thorns, poisonous plants, Venus flytraps, pool of diseased water; meanwhile Robot Shephard is far cleaner, drawing inspiration from the subgenre of Solarpunk.

Solarpunk is essentially the antithesis to cyberpunk. While cyberpunk visualises a dirty, dystopian future, where technology has overcome the natural world, solarpunk instead focuses on the symbiosis of nature and tech. Key elements include clean energy and sustainable farming - a lot of solar panels, wind turbines, and sleek white buildings with greenery growing on them. To explain what I want the game to look like, I will provide a mood board.

A lot of solarpunk images were AI-generated, so it took me a while to filter out the slop and find some more novel designs. A very solarpunk-esque architectural style is 'green buildings', in which plants are encouraged to grow on walls and ceilings. Of course, canonically in Robot Shephard it would be more unintentional, but there could be elements like biodomes and vertical farms that would have been used to provide oxygen to the facility, but have since grown out of control. Overall I think this is quite a laid-back and optimistic aesthetic, which works for the positive themes of Robot Shephard.

[nitrosodiumfmpsequel]

Robot Shephard - Ethical Research

My idea for Robot Shephard is far more relaxed of a game than Metal Jungle would be. I don't plan for it to have any violence - robots could possibly be destroyed at points, but they're not humans and it wouldn't be particularly realistic. Once again, I don't know how to include gambling, and I don't want to.

Onto accessibility, and I think that this idea might be more accessible than Metal Jungle. While a combat-focused dungeon crawler would constantly test your reflexes, I think that you would mostly be able to play Robot Shephard at your own pace. It also wouldn't contain anything particularly stressful or scary compared to Metal Jungle, which means that not only is it accessible to people of varying ability, but also age. Then, I would also stand by my tenets of having both audio and visual cues for things, not doing too much colour-coding, et cetera.

Now, diversity. Because the game is entirely robots except the player, and I don't plan to show the player throughout the game, diversity shouldn't be a massive problem. I don't think I'll give the robots too much human personality, or do any sort of racial coding (definition: Identity Coding - TV Tropes), so it wouldn't really be possible to lean into stereotypes or anything like that.

[nitrosodiumfmpsequel]

Robot Shephard - Similar Concepts Part 3: Antlions in Half Life 2

I decided to focus on the bug/ant theme, and look at the Antlion sections in Half Life 2. By this point of the game, you've been plagued by the Antlions for a good few chapters - big insects that burst out of the sand to swarm you. However, you get your hands on the Pheropods, or "Bugbait", which lets you command the Antlions as your personal army. You can squeeze the Bugbait, which makes the Antlions come to your side, and also throw them, which causes the Antlions to run to wherever they land. You can direct them to attack enemies or keep them close, but at some points you'll need to deactivate Thumpers, big piledrivers that keep the Antlions away with vibrations.

For Robot Shephard, I could spin this on its head, and have areas that only robots can enter, or buttons that only they can press if you direct them to it. The theme I want to reinforce is the cooperation between the Shephard and the robots. Neither one can escape without helping the other - the Shephard can bridge gaps or open new paths out, and the robots can operate various interfaces that the Shephard can't access.

[nitrosodiumfmpsequel]

Robot Shephard - Similar Concepts Part 2: Ants and Machine Learning

As a child, I had a book about "formica robots", which as far as I understood, were nanobots that could communicate between each other and work together to solve goals. Their name and problem-solving logic was inspired by ants, and the way that they work together as a swarm.

This has actually been a source of interest for roboticists. They've looked at replicating the efficient cooperation methods of ants, by building swarms of simple robots that can come together to accomplish goals, rather than one complex robot that could do it singlehandedly. It raises some interesting questions, and is definitely relevant for Robot Shephard, since you are getting a group of basic robots to work together and escape. Collaboration could be a key theme in the game - having pressure plates that require multiple robots to press, or using robots to climb up onto platforms that then let you do something else, kind of similar to Thomas Was Alone in that regard, but 3D.

I did also look at this research paper about directing insects to perform tasks. I'm not sure how relevant it'll be, since it's getting away from my original concept of robots a little, but it was interesting. I read that UV light can be used to direct ants (digging a tunnel was the example used in the paper), so that could perhaps be used as an in-universe way of explaining how the Shephard can direct their robot companions.

[nitrosodiumfmpsequel]

Robot Shephard - Similar Concepts Part 1: The Survivors

The Survivors is a browser game I played back in school. It's a puzzle game, and in each level you must clear a path so 40 survivors can make it to some kind of vault. This means putting bouncepads and ladders at the foot of cliffs so they can get up, drilling through walls to make passages, and covering chasms with varying sizes of bridge so they don't fall in. I would say the enjoyment comes from figuring out inventive ways to get the survivors from A to B - levels are fairly open-ended, and you are encouraged to restart if you misplace something. If this game was 3D however, or even first person, it would add another level of complexity to the gameplay.

As the protagonist of Robot Shephard, you could have the ability to place down various things to help your metal flock through, like in this game. I can't think of any first-person games that have any mechanics like this, so it could be interesting to look into further.

[nitrosodiumfmpsequel]

Game Idea 2 - Robot Shephard

This is another idea I had based off the overgrown facility aesthetic. I thought of this after looking at Duskers, the game where you control the drones and maneuver them through spacecraft. The mechanic of moving various robots around to complete tasks seemed interesting, and I think doing that with more fidelity (i.e. so you can see the environment around you) seemed like an interesting concept. The aim would be to clear a path so you can escort robots to the exit of a place, similar to the old Lemmings game but 3D.

The story concept would be that the facility had been abandoned for years, possibly centuries, and left for nature to reclaim it. It used various robots inside as a workforce for tough or dangerous jobs, but they were left behind once the people evacuated. Your task is to finally bring the robots to safety, since you are the only one that can get them through the now-damaged facility.

To save time and keep things simple, I'm going to explain my thought process of how this could work with a mindmap.

[nitrosodiumfmpsequel]

Metal Jungle - Technical Feasibility

Before I start any sort of development, I should make sure that everything I want to include in the game is possible with my current skill level.

I'll look at procedural generation first. I want to do a linear system where new rooms are streamed in as the player goes through doorways. This is essentially how we were taught how to do it in the rapid prototyping project - different room variations would be put into an Array Variable, probably named something like "Rooms", and they would be randomly selected and then spawned in at the room's doorway.

This is somewhat crude, but I think it explains my intentions with the code here. If the player overlaps with a collision component probably in the exit doorway to a room, it randomly selects a new room. If I wanted to eventually add an exit room, I could do that using another Branch, and some sort of variable that counts up.

As for shooting mechanics, I can do both hitscan/line trace for bullet weapons which hit instantly, and projectile weapons which have travel time, like grenade launchers or flamethrowers. I did line trace weapons during the Fulcrum project, and projectile weapons during both Crucible and Overdeath, so that should be fine.

One thing I am curious about, however, is how I would do a random loot table. I did experiment with this a little bit in Fulcrum, as crates would either drop health pickups or rockets when shot, but if I want to have lots of items that you can randomly pull from crates or what-have-you, it'd require a load of Branches all connected together into a statemachine.

I looked on the Unreal subreddit, and as it happened, someone was having the same problem as me. Unfortunately I couldn't access the Imgur gallery for some reason, so I couldn't exactly see their code, but I understood the problem based on what they had said. The ways of fixing it that I saw in the comments looked very complex, and used things that I hadn't experimented with before, including 'curve floats' and making new Unreal classes in C++.

I found this video, and while it is for Unreal Engine 4, I think that most of the code would still work in UE5. It's what's called a 'weighted loot table', which essentially means that some objects are more common than others. This mostly solves my problem, but I think a statemachine would still be needed to give items to the player, possibly by means of a variable. It'll probably become clear once I start development in Unreal.