While process mining and process intelligence are establishing themselves as practices that improve key business metrics, in other scenarios, they can focus entirely on the human element. Here, outcomes surpass efficiency metrics like cost-savings and profit margins, and can directly impact human lives through government systems in healthcare, justice and mental health.
Shedding light on how process intelligence can be used to facilitate better research and potentially justice reform, researchers open up about the opportunities to improve the disconnect between key systems in a large US state, working with data intelligence leader Celonis. Additionally, Celonis VP and leader for the UKI, Rupal Karia shares more about the progress of the UK’s National Healthcare System (NHS) process mining project which has already reaped quantifiable benefits.
NGO keeping Texas’ juvenile justice in check
Sharing her first-hand experience of working as a juvenile probation officer in the town of San Angelo, Texas in the late 1990s, Erin Espinosa explains to ERP Today about one of the many such cases where a ten-year-old boy, Chris, was entangled in the fragmented state of the juvenile justice and mental health systems that don’t always lead to the best outcomes.
Imagine if we’re able to process mental health and juvenile justice data to know sooner…
Espinosa, now a director of research at Evident Change, shares the story of how despite being just a minor, the child was arrested several times and would come to her probation department for misbehavior, and later assault in the classroom, which happened routinely over three years. “What we didn’t know was the other half of the story, that he was diagnosed at age eight with an early onset of Bipolar 2 disorder.” Chris’ mother also could not afford to fill in his medication prescription, so he would be weaned off his medication over the summer break and restart it when school started in autumn, further complicating the case.
This is where the disconnect between the two systems, as playing an important role in treating, rehabilitating and reintegrating youth into the community, became obvious.
“When I learned all of that information, I was finally able to intervene differently,” Espinosa says. “Chris, at 13 years old, was on his way to be committed to state prison for five years […] So that’s the scene.”
While pointing out several underlying issues and systematic inefficiencies, this case also exposed the reality of the ever-increasing ratio of children to caseworkers and managers. “Imagine if we’re able to process mental health and juvenile justice data to know sooner from a system’s approach instead of [blaming] a kid,” she adds.
This is one example of a niche where having an end-to-end view of the process can make a huge difference to human lives. This means a change from systems thinking, with separate systems (juvenile justice and mental health) that work in silos, and instead integrating the children’s information from said systems to understand what’s happening and offer better support.
To act on these particular issues, a team of researchers explored the opportunities by adding a temporal component – tracking children over time – and applying process mining into the system. “So by applying process mining and an object-centric approach, we are able to retrace the journey of these children through both the juvenile justice and the mental health care system, and combining these two, we can also see how the two systems are interacting with each other and how the handover is happening,” Arturo Castellanos, assistant professor at William & Mary, says regarding the research.
But the ultimate goal of such efforts is how to scale from one child to many more by recognizing the trajectories of children within these systems. “And by doing that, then we can take a granular level view and say “this is the individual kid, but we can also abstract that to a more general view. These are all the kids in the systems, and notice how many deviations from a typical process there are. This is where you can start enacting change by saying, ‘There’s got to be a better way, right?’” Castellanos explains, while adding that these insights can provide ideas about what needs to change in either system and the need to connect the two systems “so that we can look at the reality of each of the children in the general system”.
Process intelligence can help determine how systems perform where a child with history of arrest also has mental health issues
Castellanos describes how the teams have used process mining in partnership with Celonis to help determine how the two systems perform in those cases where a child with a history of arrest also has mental health problems – suggesting that there is a good chance that the reason for the arrest could have been mental illness that was never originally diagnosed.
For this reason, the team has come up with two KPIs. “Firstly is what we call the detention spiral. The detention is a flag which shows whether a child is more likely to come back into the system through re-offending, compared to their peers. The second is the rate of treatment accuracy. This is simply a percentage of time a child was provided the right level of healthcare for their needs,” Castellanos explains.
By combining these two indicators, the team can then gather a good idea of what can happen to these children in the long term, fully exposing that in a lot of cases where the children are not provided the right level of treatment, they go back into the system for subsequent, and potentially bigger, crimes.
Evolving mining strides in UK’s National Healthcare System
1,800 appointment cancellations were avoided monthly, saving an estimated £2.8bn
In another use case which has already seen some progress, Celonis’ VP and leader for the UKI, Rupal Karia, recently discussed the process mining efforts within the UK’s National Health Service (NHS), as the work has significantly reduced waiting times for patients.
“Unlike most of the use cases where the impact is financial, that’s not what we’re about – it’s about patient care. So it’s a completely different measurement to what naturally businesses look at,” Karia says. “In this scenario, of course, it saves money to be more efficient, but the drive is all about patients. It sounds corny, but it is improving health and therefore saving lives.”
In the past 18 months, Celonis has been working with five trusts across the UK, having reduced the waiting list from 73,000 to 67,000 patients over eight weeks, avoiding 1,800 appointment cancellations monthly, and saving an estimated £2.8bn in the first year.
The team is now looking for the next focus of this venture, aiming to expand its work in chunks of 10-20 further trusts, especially in light of the new UK government’s target to slash healthcare waiting times with sustainable solutions.
Planning the next steps for the Celonis team, Karia shares more about the trajectory of the project: “We’re now quite embedded in all five trusts and that gives me more confidence, because when you get one right, and you start seeing it work in four or five trusts, then once we get another five or ten going, we will have critical mass. It’s proven that we can do it at scale,” he says.
While the repeatability of good practice has made scalability in this sector easier, it’s also about going deeper and finding additional use cases. “There are some really complex things, like in cancer research, pharmaceuticals around vaccine deployments, as we do things around research, and we think about a different marker altogether. So where do we go next? There’s so many opportunities. We’ve got to work out where the highest impact is.”
In another step forward, the team is exploring ways Celonis can help them to be more proactive by letting them identify and address small problems before they become major issues. In this way, the approach has a good potential for being used to prevent challenges by examining a process and predicting how it can impose constraints before they materialize, something that healthcare systems in the country can greatly benefit from.
The impact of acting on process intelligence’s potential
What if we were to embed process intelligence into our
government processes?
The power of process intelligence in these cases lies in the ability to identify historical issues, clearly understand their correlation and identify potential solutions while continuously monitoring processes. This helps to track how process changes are working and make future adjustments. “We have a way to look at it from a process perspective in juvenile justice – what does that reoffender look like? Why are they getting stuck in this endless loop of reoffending? Is there a way for us to intervene before this happens?” Monica Chiarini Tremblay, professor of business at William & Mary, says, while adding “also, changing out KPIs like treatment accuracy can help us look at this object-oriented approach because there’s many ways for us to cut and slice the data.”
Approaching it from a taxpayer’s perspective, “What if we were to embed process intelligence into our government processes? As taxpayers, we have the right to know if our processes are working right,” Tremblay says. “We believe that we have a real opportunity to use process engineering and process intelligence to improve government processes.”
Using the opportunities of this pilot project in partnership with Celonis, the research team is analyzing more data from government agencies and is already demonstrating the power of using this set of data to make an impact.
When it comes to Karia and his team’s process mining journey in the public healthcare industry, they are seeing much more interest from CEOs of trusts and heads of departments seeking involvement in similar projects – a serious shift from just tech-department interest in the past.
Celonis has also recently won a contract with the UK Cabinet Office, which ensures the effective running of government. More details are to be released soon, marking the company’s entry into the UK central government, with Karia seeing the potential for more departments to benefit from the new approach at local and national levels.
As Brexit has evidently added more complexity to government structures in the UK, “there’s so much we’ve got to work out, because the reality is, we can mine any process. At the moment, we’ve started on a journey with a pretty simple use case, but I think the potential is much, much broader,” Karia says.
Going beyond the markers of traditional business wins, process intelligence is surely making its mark in public and private organizations, showing a potential for advancing business processes as we know them.
