When I saw on news on Friday that the BBC completely dropped their DMI project I was surprised, not by how much they had spent, but the fact that they had dropped everything. Over the weekend I was asked by someone outside the industry on my thoughts, and it sparked me to add my thoughts on it here. This is a conversation I’ve had a number of times with people over the years.
As part of my job (and I probably ought to make it clear at this point these are my personal opinions) I’ve been able to watch the DMI project from the distance and interested to see how it progressed. Over the years I’ve been involved in a number of projects that involve putting in digital asset management systems, and it’s interesting to see how they evolve and talk to others about their similar experiences.
As an industry television has been fairly late coming to file based working – the size of files, the connectivity required to move them and the storage required makes it a challenge, however these are all getting easier, so there is less reason why not to. In fact it’s about 6 years since I was involved with my first inter-site asset management system, so most people would assume that the BBC already had a system in place.
One thing I’ve seen over the years is what I call “my first asset management” syndrome. When the initial work is being done to scope the system the danger is that it very quickly becomes a system to solve all problems. Do you want it to store finished programmes, rushes, graphics, scripts, manage the ingest process, QC, FPA checking, approval processes, desktop viewing, logging, editing, auto-staging to playout, compliance workflows, archive, transcoding and delivery to VOD, partial restore of clips, multi-site…. the list goes on and on. Beyond that the project gets lumbered with sorting out all the business problems since “it’s a good time to do so”. Before long the list of requirements is so long you can’t see the end of it.
So the next stage is to find a system to do this for you. Many of the requirements can be served by a range of systems, however there will always be some that are not part of the standard product. So the sales team of the closest system gets to work and promises the world.
The reality of that first system then hits home – the software is installed, some of the customisation done and then it needs to be actually used. Now how well this works depends on how focused and organised the customer is and what they finally bought – if they asked for customisations that enhance the overall product that’s fine, however the danger is to get custom workflows hard coded into the system. This means that if the requirements weren’t right or the vendor didn’t understand it fully you can end up spending more money to fix the issues. Then six months down the line something else changes and it’s more money to adjust the system to cope.
The system will probably work fine for a couple of years, and then the next big shiny version comes along, or a requirement for some new big feature and then the fun really starts. This system which you’ve finally got settled and working and everyone knows how to use it suddenly is so customised that it can’t be directly upgraded to the latest version. Either you need to get the functionality re-worked to fit the new version, or find another way to do the tasks, plus you have a large migration challenge to get the media between systems.
This might all seem easily avoided, however I’ve watched a number of very large broadcasters go through this process now and all end up with systems that didn’t do quite what they expect. These days I try and stick to the philosophy of only buying a product for what it actually does and working with it (unless there is a very obvious small piece of development that definitely fits within that world).
So how does this all fit with the BBC. The reality of most companies is that money becomes the overall driver within the project. When the scope of work and customisation starts getting too big and the project overspends start to grow then that normally provides a balance to stop it getting out of control. However on a BBC budget that doesn’t appear to have been the case.
In the BBC’s case the requirements were to fit the whole BBC – through production, delivery, and archive. Early versions even managed to missed out the fact that material needed editing despite it’s size and all the thinking that had gone into it. In all iterations (pre and post Siemens) the decision seems to have been that there was nothing on the market that would suit the BBC, so they would have to write it themselves. It is this point that was probably the downfall – once you start down that route then it’s much more difficult to argue about what the system does and doesn’t do as you’re writing it yourself. I do know in the later stage of the project then there was more diligence around this, however I think this was to try and manage the scope and get something delivered.
Ultimately I think the downfall of DMI was the size – trying to handle all the requirements of both production and archive across the sheer number of sites in a custom written system was always going to be a struggle. Had the scope been reigned in – archive would have been the logical choice – and started on a smaller scale – one core site eg London – then it might have got somewhere. Trying to build all features on all sites is ultimately what killed it.
So where do they go now? To my mind the BBC should look for a single archive system based in a single site that other areas can view and deliver into. There are systems off the shelf which could provide this right now and would get them many of the benefits of file delivery. In the mean time many of the productions and production sites have probably adopted file based working, so work with what they are already doing.
At this point the BBC has some real catching up to do with the rest of the industry. They need the ability to receive a file from production, send it on to other places and archive it for posterity. However what they’ve been through is not unique, but it’s on a scale only the BBC could achieve.