This article was first published in the Spring 2014 issue of Prime

The analogue world is one Grandpa would find quite reassuring and comfortable. The phones are old-school rotary, the clocks have hour and minute hands clicking away their endless circles, and music comes from record players. Throw in some Pabst Blue Ribbon beer, and you have a perfect retro-cool setting.

Manufacturing environments are increasingly modern and high tech – but they embrace analogue as well. Analogue devices work in a linear manner (as opposed to the on/off aspect of digital) and can best capture the continuous nature of voltage, resistance, pressure and other common factory floor measurements. These analogue devices fill the factory floor and send out data with every cycle. This data, however, has a sad and pathetic life with most of it being utterly unused past the point of creation.

The growing interest in intelligent machines – whether called machine to machine, the Industrial Internet, or Internet of Things – shows that this data could and should have a life that is perhaps longer and certainly less pathetic. There is good value in extracting business insights from data that is available for free. Having a strategy to collect, store and analyse analogue data can increase return on investment (ROI) for assets and resources and ultimately lead to better business outcomes. Benefits include:

• Reduced unplanned downtime
• Reduced maintenance cost
• Improved reliability
• Increased safety
• Predict and prevent failures
• Monetise data.

Building common ground
Converting disparate piles of analogue data into something useful requires collaboration between experts in operational technology – the manufacturing people – and the IT department. This is a significant barrier as IT people and manufacturing people are like natural enemies in the wild. IT folks live in a world of trouble tickets and roadmaps, whereas manufacturing lives on the front lines, battling the daily fires to keep production flowing. But each brings vital knowledge and incentives to see ROI projects through completion.

Manufacturing people understand how data is generated on a factory floor. They know what the sensors are measuring and bring some initial ‘gut feel’ as to what streams are most critical to analyse. More critically, manufacturing folks are in a position to champion projects that improve quality, efficiency and cost – they’re graded on it every day – and can supply the urgency and drive to get projects approved.

IT people can help transform the raw data into something quite useful. They can use local computing resources to collect the sensor data and convert analogue outputs to digital. They can set up the algorithms to aggregate peaks and valleys in the data streams and deliver near real-time alerts on key metrics, such as quality or production goals. This is an innovation hot spot, since the value of catching issues early is high. For larger projects, such as predictive analytics or communication between machines on different standards, IT people can bring out the big iron and pull together solutions combining the cloud, software and networking. Cloud technologies are a solution to the lack of standardisation in operational technology, as it is technically much easier to have devices talk through the cloud than it is to have them talk directly to each other. Using rapidly evolving software, IT professionals can integrate data streams from sensors, production systems, warranty databases and even supplier sources. The software can pick up on patterns that lead to certain outcomes, and this learning over time is the heart and soul of predictive analytics.

Maintenance scheduling is a common use case for predictive analytics. Instead of replacing parts or equipment on a fixed schedule, risk models are created and parts are only replaced when a pre-determined degree of wear and tear is detected. We see this in oil changes in newer cars – instead of changing oil every 3,000 miles, we wait until the oil life indicator hits a certain level. This is more accurate and for most people will reduce the amount of time and money spent maintaining their vehicle.

In the end, the real barriers to creating intelligence from analogue devices are not the technologies but rather the lines on an organisational chart separating manufacturing from IT. Getting natural enemies to collaborate will take organisational commitment, a clear roadmap and generous servings of coffee and doughnuts. Here is a starting workplan:

1. Set a clear objective, with shared accountability
This mission should come from a higher level than the functional leadership of manufacturing or IT, and should ensure that both functions have skin in the game.

2. Learn together
Take advantage of the whitespace for both functions, and have functional leads learn together through conferences, webinars, or meetings with consultants. Spending time together in a low-stress environment can help build natural rapport.

3. Target simple wins at first
Gain momentum through tackling the low hanging fruit of simple yet impactful projects. Relationships will grow through joint problem solving, and achievements will build confidence to tackle bigger problems

4. Celebrate the wins
Reinforce the mission through widespread communication and celebration of wins. This also attracts the attention of others who will become eager to lend their time and talents to a winning team.

The factory floor may be filled with analogue devices, but as these machines become more intelligent, it is no longer your grandfather’s factory.

Kirsten Billhardt is the global manufacturing strategist at Dell

Share this story