Application of AI enhances field service management for 5G, IoT and beyond
As communications service providers (CSPs) gear up for 5G and look to support IoT businesses as a new business line of their own, field service engineers will be under more pressure than ever. However, artificial intelligence (AI) and machine learning are being applied to the discipline of field service management to optimise field service performance through a blend of intelligent routing, remote assistance and performance analysis. VanillaPlus managing editor George Malim talks to Paul Whitelam, the senior vice president of global marketing at ClickSoftware, to catch a glimpse of the field service management of the future.
George Malim: Field service provision is a substantial cost and management burden for CSPs and it’s likely to become more of a challenge as mobile network equipment is more densely deployed and CSPs include devices for indoor coverage and IoT-enabled equipment in their offerings. How are CSPs optimising their mobile workforces?
Paul Whitelam: We’re making sure that mobile workforces are as efficient and effective as possible. Our intelligent, AI-driven core ensure people with the right skills, parts and knowhow are in place at the right time but that’s only scheduling optimisation which ensures optimised routing and minimises things like time spent in traffic. Beyond that, we focus on optimising the day of service in which there might be 1,000 engineers doing 10,000 jobs. We ensure that you are set up for success in this, using predictive analytics to identify what the demands of each job are and to ensure people and necessary parts are in the right place.
GM: How granular do you get? Is it possible to get down to the level of an individual engineer in a specific city suburb?
PW: Yes, it’s very much down to that detail. An engineer’s service day is very much driven by how long it takes them to do a certain task and how long it takes them to get to a certain job. For example, one engineer might be an expert at deploying a specific type of router and can do the task in 30 minutes when others take 40 minutes. However, that advantage could be negated if they are routed in a way that means they spend more time in traffic between jobs. The value here is that we ensure that resource utilisation levels are high.
We do that on the travel side by using predictive travel information. We know from Google data what the likely time needed from point A to point B is at a specific time so when the day of service is planned enormous mathematics can be used to juggle all the permutations to come up with an optimal approach.
This granularity is where we are different. Other field service management providers generalise and say that on average in an eight hour day an engineer will be able to do six jobs. The problem with that approach is that, on average, you will be wrong.
Another aspect we bring in is the job duration. You can come up with specific engineer profiles that help to ensure tasks they are strongest at are assigned to them and there are other parameters such as weather conditions, the brand of router being deployed or whether it will be necessary to wait to gain access to infrastructure at a business location, for example.
GM: Although the savings and operational efficiency gains are substantial they pale into insignificance against CSPs’ costs of network build so how do you get their attention?
PW: It’s certainly straightforward to explain but this isn’t a project these organisations do every year. It is a major deployment and we recognise that it’s a big investment but it has big return on investment attached. We’re at Deutsche Telekom, Oi and we have 70 leading global CSPs that are getting the benefits from our system. We work with them to construct business cases that address their priorities. Those might be improved first time fix rates or reduced windscreen time for engineers, both of which can be quantified, ultimately in cash terms.
A key question to consider is if you can reduce each engineer call by ten minutes per job what does that do to your business?
GM: What about the increased amount of equipment that needs service and maintenance with the arrival of 5G and the scaling up of IoT?
PW: We do see both of these happening. 5G certainly means there will be more projects and for us it matches very week with concerns CSPs have about the amount of training that will be required and how they will meet that need. One way to do this is to use personnel more effectively at the point of service. This could be achieved by focusing skills on a few people that are running round all over the place. You could place experts centrally that can support engineers on the road and take advantage of enhanced capability to provide remote assistance.
IoT itself is interesting and CSPs could carve out additional roles here in supporting the infrastructure required. The complexity IoT brings is welcome for us. People are investing in real-time visibility into their assets but there’s no value in knowing a machine part needs replacing if you can’t ensure your field service management can allocate an engineer to perform the replacement. Out system can look at the best possible response time to enable emergency replacement to be performed and then automatically back fill the gap created by doing that in the day of service.
IoT is already a huge driver and we believe it will drive more as it grows.