Starting something moving takes more force than keeping it moving
Static against dynamic friction, and where it causes trouble.
Starting something moving takes more force than keeping it moving because the microscopic contact points between two stationary surfaces have time to settle and interlock more fully while at rest, giving static friction a head start over the sliding friction that takes over the instant motion begins.
A drawer that sticks, then lurches
An old wooden dresser drawer swollen slightly with damp needs a real, deliberate push to get moving. It resists firmly right up until it gives way, then slides open much faster and more easily than the initial push seemed to demand, sometimes lurching far enough to catch a person off balance. That lurch is the gap between static and sliding friction made obvious. The drawer's resistance drops sharply the instant it moves, and a hand still pushing with the force that overcame the initial stubbornness is now applying far more than the sliding drawer needs.
A drawer opened every day rarely develops the same stubbornness, because regular use never gives its contact points a long, undisturbed rest to settle into. The drawer that finally sticks hard enough to lurch is usually the one that sat closed the longest.
Why contact points settle at rest
Two surfaces resting against each other touch only at a scattering of microscopic high points. Given time at rest, those high points settle further into each other, deforming slightly and increasing the real area in contact, which raises the force needed to break the contact and start sliding. Once sliding is underway, the contact points break and reform far too quickly to settle in the same way, so sliding friction is generally lower than the static friction that had to be overcome first. The gap exists for essentially every ordinary pair of dry surfaces. The figures printed in many physics textbooks put dry steel on steel at a starting coefficient of about 0.74 against a sliding one of about 0.57, so it takes roughly thirty percent more force to set the steel moving than to keep it going.
Static friction is also not a single fixed value for a given pair of surfaces. It can grow slightly the longer the surfaces have rested together, which is why a machine left standing overnight can feel a little stiffer to move the next morning, at the same joints and the same load, than it did the evening before.
Stick-slip in slow, precise motion
A mechanism driven slowly and steadily against this gap can fall into a repeating pattern called stick-slip. It builds up force while stuck, breaks free once that force exceeds the static limit, accelerates briefly because the lower sliding friction cannot absorb the force that was needed to break free, then slows and sticks again once the excess is used up. The size of the gap sets the size of the jolt on every restart. For a machine tool's slide, a robot joint or a positioning stage, stick-slip turns an intended smooth crawl into a series of small unpredictable jumps, each landing the mechanism somewhere slightly different from where smooth motion would have put it.
The same cycle, run hundreds of times a second, is familiar as sound. A squeaking door hinge, chalk screeching on a board and a violin bow drawn across a string are all stick-slip. The rosin on a bow grips the string and drags it sideways until the string's own tension breaks it free, it snaps back, and the bow grips it again, which in a violin is the whole point and in a machine slide is a fault.
Speed changes the problem too. A mechanism crawling very slowly spends long stretches stationary between slips, giving static friction repeated chances to rebuild, while the same mechanism driven a little faster spends less time at rest and can ride through the gap far more smoothly than the gentle, slow approach that looks like the careful choice.
Narrowing the gap, or never stopping
Avoiding stick-slip means either narrowing the gap between static and sliding friction directly, with a well-chosen lubricant or surface treatment, or avoiding the conditions that let static friction build, keeping a mechanism in continuous gentle motion instead of repeatedly stopping and restarting it. A surface that never settles at rest never builds the larger static friction that causes the next jolt. Rolling contact, covered next in this set, sidesteps the problem in yet another way, since it barely experiences the settling effect at all.
A control system driving a mechanism through the threshold where static friction gives way can also be tuned with this behaviour in mind, anticipating the sudden drop in resistance before the mechanism jumps. A loop tuned that way handles stick-slip gracefully, where one that only reacts is caught out on every restart.
The size of the gap varies considerably between material pairs and surface conditions, and the only reliable way to know it for a particular pairing is to measure it. Some combinations show barely any gap while others show a very large one. Choosing a pairing with a small gap is itself a legitimate design lever wherever stick-slip is a real risk.