NOTE: This post has been updated to include response from Pascal Mueller Director of Esri R&D Zurich and the creator of Esri CityEngine, scroll to the end to see his response.
CityEngine 2018.0 beta has been out for a while now and is pretty much over, I’ve been lucky enough to have been using it in anger for a bit too. Building on the CityEngine 2017.1 release there are some small improvements, new CGA code as well as a couple of new features, but essentially this feels like a minor release which is not a bad thing. Just because it’s minor doesn’t mean it doesn’t feel like progress , it will be well worth the update!
Firstly a confession, the title is a bit of a click-bait (hey if respectable news organisations can try it why can’t I?) I can’t show you a sneak peak this time or really talk about specifics, well because of um beta stuff and all that…. (last time I had a special agreement). What I can tell you is that if you want to participate in the future CityEngine betas you can sign-up here https://earlyadopter.esri.com . If you like CityEngine and are interested in where it’s going and of course helping out the team at Esri then I thoroughly recommend you sign up!
I can assure you that some bug fixes have happened and a couple of new features added which complement the analysis side of things.
Planning, designing and development must take the full picture into account. Create, visualize and share in 3D to make better designs and present your work more effectively. Source: Esri
This I guess poses a question for Esri, does it continue CityEngine long term with so many products that seem to tick the urban space box? As someone who has trained many clients in CityEngine and offers consultancy services over the last 5 years I certainly have an interest in its future! There are big questions of whether it will be just consumed by ArcGIS Pro or placed in some cloud container and interacted with via ArcGIS Online.
I think (and I have no insider knowledge here) the answer is all of these things will happen, but also that CityEngine will continue to be a standalone product for the foreseeable future. Esri is in the movie industry now and I don’t think it will want to leave that, but also CityEngine is a great design tool that fits Jack’s interest in urban and geodesign.
“is a simple-to-use 3D city editing and visualization tool” – “Build Flexible Scenarios Faster” – “Create Realistic Context”
The real worry for me is what form will CityEngine evolve into. The marketing literature still talks of easy operation and high quality realistic outputs. Sorry but it’s not easy to learn or use (from a typical planners perspective) and the high quality outputs don’t come instantly (unless you like Redlands building typologies!).
CityEngine’s great power is it’s flexibility of format support and usage. But it has fundamental technical issues/challenges with terrain and roads that need addressing. Esri also needs to decide whether this is a tool for the drag and drop users among us, or some highly technical development environment for urban planning.
One glimmer of light and a direction of travel is the ESRI.lib folder, promisingly for new users it shows a path of drag and drop produce something now. Unfortunately this seems to have gone a little stale, anyone who does CGA coding in CityEngine knows organising and keeping up to date rule files is difficult especially if projects vary like ours do. The ESRI.lib has some great rules I use over and over again, roads and trees, in fact the more generic the better! Give me more of those tree assets! These smaller rules enable a greater freedom and I can write rules that work with these easily because I know they are installed with every CityEngine workspace.
In conclusion, and I’m not good at concluding thoughts so forgive me, CityEngine is here to stay but has some challenges to overcome. Like all software it has to evolve as it’s users ask more of it. Currently I worry that CityEngine as a tool is too technical for widespread adoption this means that businesses will be reluctant to invest in it if only a hand full of users (CityEngine professionals) are out there. On the plus side CityEngine professionals like myself will probably have some consultancy work coming our way!
Update and response from Esri Zurich’s R&D Director
So I posted the link to this blog on LinkedIn and amazingly Pascal Muller (read a post I did called Life Changer to understand why I am honoured to get this) very kindly responded. I have got his permission to post those comments here, these have been copied directly and no editing has been done although I have tried to keep the paragraph breaks. What I now need to do is a follow up post to contemplating this response!
Thanks Elliot, really great read!
I could answer some of the raised questions : ) hmm, maybe we should do an ‘ask me anything session’ here…
(1) You are correct, not a ton of new features in this release. Reason is that we changed our release cycle this year and are releasing now 3 months earlier. As a result the time between CE 2017.1 and CE 2018.0 was much shorter. CE 2018.0 comes out next week and you can expect CE 2018.1 in September instead of November.
(2) Yes, CE is here to stay. There are no plans to discontinue it. In contrary: maybe you have heard about our new product initiative ArcGIS Urban which is basically a streamlined urban planning platform (on top of AGO) for the planning departments of cities. It will feature cool web apps/interfaces and it will also work great with CE. It’s all developed under the same roof here (also includes the 3D JS API 4.x team btw). In the ArcGIS Urban context, CE will (still) be required for the more advanced workflows such as for example Devin’s site plans, Bruno’s Masterplanner, or the various greenfield mega city projects of the HOKs and F+Ps.
ohh, and of course Blade Runner 3 will need these off-world cities Batty was talking about : )
Besides all this, CE continues to be one of Esri’s development platforms where bleeding edge technology gets applied before it goes mainstream, see for example the push on game engines (more about that later…).
Pretty exciting stuff and I am extremely happy that all the pieces fall into place finally, but we know that we have to keep working hard and improve things. Huge thanks for your continuous support and wise long-term thinking.
This all started a while back with an interest in American houses, specifically the ‘Craftsman’ style homes which you could buy from places like Sears of all places…. As an Englishman I’m not fully versed on all North American architecture but this idea of picking elements of a house from a catalogue seemed like something I should look at doing in Esri CityEngine.
So that’s what I did. In classic CityEngine professional style my first attempts got very very complicated as I added more detail (you’ve seen the nice renders of these for a while now on this blog and even the banner here, I really really like them). The second attempt I produced an all singing all dancing rule file to create thousands of different ‘Craftsman’ style house typologies. The trouble is, without a big PC and a good understanding of CityEngine these were only going to be usable by those of us who sort of know CityEngine in a professional capacity. This is always a problem I have with CityEngine much like Alice I often go deep down a rabbit hole and get lost in the wondrous and slightly crazy detail (it’s not a bad thing just a bit distracting!).
Third time lucky?
The end result on my third attempt is something very much simpler and easier to understand and something that can be wrapped as a Rule Package. As a Rule Package it can also be used in ArcGISPro too. I’ve initially conceived of this rule file as primarily use on auto-generated Lots in Esri CityEngine (those are created from centre lines of roads forming blocks), but I can make it work on points and footprints too (just not on the first release). In CityEngine you have the nice ability to use the handles feature to interact with the model without having to muck around with the attributes in the inspector. Here I am finding it quite tricky, which attributes are important to have as ‘handles’? If you do everything the model becomes cluttered, so I am going with a ‘less is more’ approach to see how it works.
Okay enough already! When is it released, and how much will it cost?!
I’ve got a beta trial coming for this which I hope some lucky few will help me iron out the kinks. Then I’ll look at selling it, my thinking is Rule Packages get sold cheaply (less than £50 probably) but if you want the source code naturally you have to pay more, and it’s this price I’m struggling with. perhaps others can suggest an approach?
Well it had to happen eventually, we’ve got our first British Cartographic Society 3D Special Interest Group (or 3DGBCS) meetup coming at the end of March, hosted at the Ordnance Survey offices. Nicholas Duggan has been leading this and will have finalised the details shortly, be sure to keep an eye out for it on social media as well as here.
Now on to more news: we’ve had tremendous support in setting up the 3D Group and we want too extend this. As such we have setup an award specifically for 3D cartography of any sort from any industry or profession. This bit is important really as the term 3D can be all encompassing and we didn’t want to limit who entered, you don’t even need to be a member of the BCS. In our work we’ve seen representations of the world in 3D from many industries just look at the entertainment industry for a wide range of 3D technologies and let’s be honest mapping/cartographic techniques. In urban planning and architecture 3D representations of the world around us provide important context for proposals. With smart cities a 3D basemap is considered integral to the bringing all this city data together.
3D Cartography Award 2018
The first annual 3D Cartography Award Sponsored by GD3D® the 3D geospatial brand from Garsdale Design, is a new exciting award open to everyone in any industry creating interesting, informative, exciting 3D cartography (real or imagined) using any technique and/or medium!
Currently we find 3D representations all around us, whether it is a web map, a survey plan, a planning visualisation, or even a computer game. This can come in many forms from a simple isometric drawing through to full haptic virtual reality.
We don’t care what industry you are in or what software you use, whether you are a surveyor, cartographer, GIS user, artist, engineer, data scientist, or other, we just want to see your amazing 3D representations and hope the entries will challenge everyone’s perception of what 3D cartography is!
There are many interpretations of what 3D cartography is, so we don’t want to limit entries. We propose an award for an overall winner based on communication of the intended message, legibility, simplicity, visual impact, and composition. There will also be commended awards for those we see as having merit in particular areas, science, statistics, visualisation, urban, natural environment, fictional, and other.
Entries will be considered by a panel of judges, appointed by the GD3D® team at Garsdale Design and the BCS Awards Committee. The panel will include a range of people from different areas of expertise in the 3D data industry. The panel will judge the quality and design of the map in relation to the purpose for which the map was produced.
The winning entry will be announced at the BCS-SoC Conference in September 2018. The award comprises a crystal trophy to be retained by the winner and a certificate. The winning entry will be put forward for the BCS Award. Those commended will receive a certificate.
All entries will be exhibited at the BCS-SoC Conference and the winners will be published in The Cartographic Journal and on the BCS website following the Awards Ceremony.
We don’t want to limit the entries but as guidance below is an example list of entry types. This list is not exhaustive, and the judging panel will consider other formats as appropriate:
VR formats compatible with HoloLens, HTC Vive, Apple & Android
In CityEngine you can use a feature called ‘handles’ to add visual controls to your models. This means you don’t have to interact with specific numbers or values in the inspector. It also means many more people can use your rules as they were intended. The real issue here is how much control do you give you users via this method?
As you can see there is a balance to be had between being highly configurable and being too configurable (i.e. too many attributes to muck around with!).
Many people have reached out to me asking whether this rule file will be available for sale sometime. The answer is probably! Drop me a message if you’re interested and I’ll let you know how it goes. Or keep an eye on twitter, LinkedIn or this blog for more updates. I suspect I will have to stop sometime.
Some of our latest CityEngine work is looking at the high street and in particular here at home the UK high street. Commercial buildings in a typical UK town are a mixed bag of traditional older buildings with some often badly maintained concrete buildings and the odd brick built modern monster designed and built in the 1980s. More recent buildings like glass a lot … We’ve been creating rules to describe building frontages, not all are pretty but that’s kind of the point!
Further to my CityEngine Quick Render post I thought I’d put through some of our GD3D buildings (sourced from CyberCity3D and processed for use on the ArcGIS platform sold via ArcGIS Marketplace) through the same process. I think they look quite nice! My next task will be to start showing more than pretty renders of buildings, I hope to add some metrics in and the do more nice imagery.
All this rendering of 3D models whilst not new to me is something I do very rarely, other people can do better but as with everything it is nice having some skills ‘in-house’!
I’ve been playing with my new ‘Craftsman’ style CityEngine rules in other workflows and pipelines. The one I’m working through at the moment is some experimental renders, I know there are a lot of issues here but I’m not yet interested in spending hours working and tweaking my rendering program to get ‘just’ the right level of detail and dirt (there are better professionals than I for that). It’s still amazing to me that these are proper GIS based models (they mostly designed to work with GIS data in ArcGISPro) with real dymanic metrics being now used to produce something that looks quite good (well I think it does).
I like how these renders make my craftsman style models look like something you may print out via a 3D printer. Okay it I know it looks toy-like, and is somewhat dated in terms of visual renders etc.. (I’m not a entertainment industry/graphic artist professional) but I still like it. Also apologies for over use of Depth of Field effect….