Crispy Mountain’s Keyline MIS is typical of a new breed built around open applications, standard interfaces and calculations in the cloud.
Three years ago one of the most talked about MIS systems today did not exist. The company behind the Keyline MIS is German software house Crispy Mountain and it was wrestling with improving logistics systems.
Then came the idea about porting this experience into helping the automation of the printing industry. In Germany and across Europe that is a growing user base from small shops to the 300-seat system at Livonia, the largest book printer in Latvia. In the UK, Paul Warren spotted the potential and Cloud to Print has been selling the technology to printers in the UK. It took a stand at the Print Show where Colourfast was announced as the first UK customer, and a second deal was completed on the final day at the C2 Group was the second. More will follow.
Keyline is still a young platform that is not yet 100% comprehensive. There is no ganging function for example, but many MIS systems that have been around for more than 30 years do not offer a fully fledged ganging function. There is enough in place already for it to be worth a look.
Joint managing director Christian Weyer is cut from the mould of tech entrepreneurs. He is a neuro scientist by training, who wears his knowledge of the workings of the human brain lightly, when introducing the software. He runs through the way that technology is seen, at first in terms of the technology it threatens before shedding those clothes to emerge as a paradigm changing technology.
Thus the desktop computer sat in an office to replace a typewriter – hence the paperless office predictions – before coming of age. In terms of the personal computer this was the introduction of iPhone. “This was the computer as it should be used,” he says.
By implication many of the MIS systems in use in print still belong to the pre-iPhone era. Keyline doesn’t. It is hosted in the cloud, giving the developer the ability to update the software several times a day without impacting user data. It is built around open APIs, enabling straightforward connectivity to third party applications. That to Salesforce was completed within a few hours.
This is not the lengthy process that has handicapped the widespread adoption of JDF (a format set in the pre iPhone age). Data is presented in a highly visual way, making it easy to understand where a job is.
Most importantly though, for Weyer, the arrival of Keyline is timely. “Thirty years ago, the industry was characterised by large orders and slow turnarounds. Over the last 20 years runs have been getting shorter and shorter and printers have to handle many more orders but without increasing the overall administration costs. In that period there have been numerous improvements in production technology, but the business has not been automated because printers are using the same software as they did 30 years ago.”
The connectivity of Keyline and the systematic approach that underpins its technology is the response. “Standardisation is needed, not of the product, but of the process,” Weyer says.
Standardisation also requires agreed semantics, what he described as a taxonomy, a structure to describe any job which becomes more precise the deeper into the description that you get. “Semantics is the foundation for automation, it is what is needed for a computer to interpret the data.” For Keyline this means a consistent way to enter data, and to build the templates that define the processes, resources and materials needed for any job.
“It’s hard work and takes time, but once you have gone through this, you can automate pretty easily,” says Weyer.
As the job proceeds, the status is changed by adding a simple adjective. Thus ‘plates’ become ‘exposed plates’ and a resource that is available to the next step in the process that would not be available until being qualified as ‘exposed’. By parsing the stage that each job is at and what is entailed for each job, it becomes possible to decide the best method to produce each job, and when to do so, batching by job type for example.
There is allowance for the non productive time, that needed for cleaning a coating unit for example, which takes a press offline, and there is appreciation too for time that each task might take: lamination taking longer than folding the same number of sheets.
A very experienced production manager might understand this intuitively, assigning T cards to the planning board as jobs are booked in. But the flow today is too fast and the rules about how long jobs take and what equipment is used to match the job definition needs to be distributed across the business.
These are the rules that can be defined by the user and which are used to build the production paths. These can be straightforward says Weyer: “If the paper specified goes beyond 300gsm, then use this press, for example. We try to get printers to use the simplest formula that is possible,” he says. In Germany this can be a challenge. No printer has yet suggested that relative humidity becomes a variable that needs to be understood, but it is only a matter to time,” he jokes.
Statistics comparing the schedule against the reality, the costs compared to the estimate, are retained for comparison and to adjust for the next comparable job. It is not artificial intelligence, Weyer insists, but application of smart algorithms. “It’s not artificial and it’s not intelligent. Our competitors might claim this is artificial intelligence but it’s just algorithms, we can call it machine learning.”
Most tasks that a printers handles are repetitive “90% don’t need to be intelligent” he says. “We want to free people from doing the stupid tasks to allow them to do the intelligent work.”
This opens the way to automation, both of pricing and especially of scheduling. During a presentation at the Print Show, he asked printers present to raise their hand if the schedule planned at the start of a day was completed exactly as conceived. Nobody did.
With feedback from direct interfaces or from iPads if a direct interface is not available to enter simple data (job at makeready, running, stopped or finished) the software can adjust the schedule on the fly accommodating rush jobs, or those where artwork has not been signed off or where the paper has not yet arrived. “Our MIS looks at planning every five minutes,” he says.
Not even the most obsessed lean management maven can do this, let alone during the turmoil of a typical day of production with the pressure to deliver overnight. This is the logistics problem where Crispy Mountain cut its teeth, where the cloud provides the ideal vantage point to monitor everything that is going on in a print business.
Crispy Mountain secured its first order at the Print Show in September, signing the deal with Colourfast at the NEC. There was a constant flow of demonstrations during the show, with a second order received at the end of the event.
Story 1 of 2
Crispy Mountain has instigated an App Store for users of its Keyline MIS to extend the system through other applications that link to the MIS via APIs. The developer plans to add links to production machinery through APIs making automation easier, one to Muller Martini already exists having been developed for one of Keyline largest users. Links to Web2print applications are underway to automate the processing of incoming jobs.
Story 2 of 2