Batch on Flow: The Physics of Lean Throughput

 alt text: Increase productivity by limiting Work in Progress (WIP)

Batch on Flow

My name is Bob Batcheler, but my friends and colleagues call me Batch. I am new at LeanKit. When I started here, I expected the ritual hazing about my last name, mostly having to do with my marital status, but LeanKit broke the mold.

Early on, either @leankitjon or @indomitablehef (LeanKit’s COO and CEO, respectively) commented that I may have to change my nickname from Batch to Flow — definitely an “inside Lean” joke. The more I thought about it, I realized it would be a great topic for my first blog post. So, here you go — Batch on flow.

Boost Productivity: Introduce Slack

When I was introduced to Lean, one of the first challenges I experienced was that so much of Lean seemed counterintuitive. This is because we are selectively programmed to pay attention to things that fit our existing mental models — otherwise known as cognitive selection bias.

One such bias tells us that if we want to produce more, we have to work harder. But this isn’t the case; the key metric isn’t effort, it’s results — throughput. To increase the productivity of a system running at 100%, we have to decrease the work in progress, or WIP. In Lean terms, we have to introduce slack (this slack, not this one).

Several writers, including our own Julia Wester in her blog post on slack, appropriately use a rush hour highway to illustrate this concept. We all know that having too many vehicles on the road creates traffic jams; a highway at capacity is a highway at a standstill. The same is true for teams — when a system is always at or near capacity, it introduces more friction. The result? Everyone slows down.

alt text: Reduce friction and maximize throughput by introducing slack

Capacity and Flow

As a civil engineer by training, the highway analogy resonates strongly with me. But that’s not the only civil engineering example that supports this principle.

Think of the culvert under a roadway, where water flows through by gravity. Interestingly enough, the pipe will carry more water when it’s flowing at 90% of the diameter than when it’s flowing full.

The physics behind that phenomenon applies to our discussion on WIP: When water flows at 100% capacity, the roof of the pipe creates additional friction, reducing the throughput of the pipe. The same is true of any process — as a system approaches its capacity, increased “friction” reduces throughput faster than it increases utilization. In simpler terms: The more you try to do, the harder it is to get anything done.

The Bottom Line

The laws of physics make it easier for us to accept the truth that evidence clearly shows — if you want to achieve flow, you have to introduce slack.

Bob Batcheler

Bob Batcheler is no stranger to Lean management principles, as co-founder of Newforma, Inc. and now in his role as Vice President of New Market Development at LeanKit. He uses his wealth of experience to develop strategies to address the critical business issues associated with Lean transformation in new markets. Follow him on Twitter @rpbatchAEC.

5 thoughts on “Batch on Flow: The Physics of Lean Throughput

  1. So, Jason, to clarify – is your question, “How do you define the optimal amount of slack to introduce into a constrained system for any particular case?” If so, there is a great rule of thumb to apply to these cases, proposed by @DReinertsen and others. Identify the current amount of Work In Process (WIP) in your process, then implement a WIP limit that reduces WIP by 1/3. Run your process, respecting that WIP limit, for some reasonable period of time and then evaluate the results. Reduce the WIP limit by 1/3 again, run your process and evaluate the results. “Lather, rinse, and repeat” until you identify the WIP limit that maximizes throughput. The resulting WIP limit represents the optimal amount of slack to introduce into the system.

    A final note – one of the common, early wins in a new implementation of LeanKit is that our customers generally have no idea of the amount of WIP that exists in their work processes prior to implementing LeanKit. When they visualize their process via LeanKit, the amount of WIP becomes apparent, allowing them to establish the WIP limits suggested above. Limiting WIP is generally the fastest, least disruptive, and lowest risk way to improve the throughput of a constrained process.


Leave a Reply