Limits: The Ultimate Get-Out-of-Jail-Free Card

Limits put calculus on a sound foundation. Whenever the math looks shaky, a careful limit is often the move that restores order. Infinity? Take a limit. A function misbehaves? Take one-sided limits. An integral runs off to forever? Take a limit. A formula looks illegal until the last second? Take the limit carefully and see what survives.

Limits are how calculus keeps its promises.

Where this principle is taught and reinforced: Lectures 01 to 02 introduce limits through Archimedes' approximation principle and limits at infinity. Lectures 06 to 08 develop one-sided limits, continuity, and the Squeeze Theorem. Lecture 16 uses limits as the "Get Out of Jail Free Card" for improper integrals. Lectures 21 to 22 return to limits in the setting of Laplace transforms and signals.

Archimedes' Approximation Principle: Trap the Quantity

Archimedes taught us one of the deepest moves in mathematics: if you cannot compute a quantity directly, trap it between quantities you can compute. Build a lower estimate. Build an upper estimate. Tighten both. If the two estimates close in on the same value, the unknown quantity has nowhere left to go.

Today, one version of this idea is called the Squeeze Theorem, or Sandwich Theorem. But the principle is much bigger. It is how calculus defines and computes areas under general curves, volumes, arc lengths, moments of inertia, and other quantities the Greeks could not even dream of.

This is not a trick. This is the essence of calculus.

Where this principle is taught and reinforced: Lecture 01 introduces Archimedes' computation of π by upper and lower polygonal estimates and uses the same idea to begin area under a curve. Lectures 02 to 04 turn this into limits and Riemann lower and upper sums. Lecture 08 gives the modern Squeeze Theorem and shows how upper and lower bounds rescue difficult examples. Lectures 16 to 17 use comparison ideas for improper integrals. Lecture 27 returns to bounding, approximation, and numerical computation in larger engineering problems.

Understanding the Whole from the Sum of Its Simplest Parts: The Amazing Power of the Rectangle

Engineers often understand a small piece of a system long before they understand the entire system. Calculus exploits this fact. Break a complicated object into infinitesimally small pieces, write a simple formula for one piece, and then add the pieces together. As the pieces become vanishingly small, the approximation becomes exact.

This idea appears throughout engineering. A robot link is not a point mass. It has length, shape, density, center of mass, kinetic energy, potential energy, and moment of inertia. Rather than analyzing the entire link at once, we analyze a tiny piece and then sum the contributions of all the pieces.

The rectangle is the first and most important example of this idea. By slicing a region into thin rectangles and adding their areas, we compute the area under a curve. The same principle later gives us mass from density, center of mass from distributed mass, energy from power, probability from probability density, and moment of inertia from mass distributed throughout an object.

If you truly understand this principle, many intimidating formulas become less mysterious. They are simply different ways of understanding the whole from the sum of its simplest parts.

Where this principle is taught and reinforced: Lecture 01 begins with the "Amazing Power of the Rectangle." Lectures 04 to 05 define and develop the Riemann integral. Lecture 06 applies differentials and integrals to links of a robot with distributed mass. Lectures 12 to 14 use rectangles, disks, washers, shells, path length, energy, and Lagrangian modeling. Lectures 15 to 16 connect accumulated quantities with antiderivatives and improper integrals. Lecture 27 returns to moment of inertia and numerical integration.

Linear Approximation: Make the Nonlinear Usable

The world is nonlinear. Linear algebra is where engineers have power.

Linear approximation is the bridge. Zoom in close enough, and many nonlinear functions behave almost like linear ones. That single idea gives us tangent lines, derivatives, differentials, Taylor approximations, Jacobians, Newton's method, robot linearization, and gradient descent.

Without linear approximation, one of the mathematical engines behind modern optimization and AI would not run.

Where this principle is taught and reinforced: Lecture 09 introduces the derivative and the centrality of local linear approximation. Lectures 10 to 11 develop Taylor approximation, partial derivatives, and Jacobians. Lectures 12 to 13 use derivatives and gradients in optimization and engineering applications. Lectures 19, 25, and 26 use Jacobian linearization to replace nonlinear ODEs and robot equations by linear models that can be analyzed and controlled.

The Fundamental Theorems, or Speedometer-Odometer Duality: Accumulated Rates Build Totals, Differences of Totals Reveal Rates.

Your car knows calculus! What?

Differentiation tells us the instantaneous rate of change. Integration tells us the accumulated total. The Fundamental Theorems of Calculus reveal that these are not separate kingdoms.

Velocity accumulates into position. Acceleration accumulates into velocity. Density accumulates into mass. Power accumulates into energy.

And, under the right conditions, accumulated totals reveal the rates that created them.

This is the hinge of calculus: local change and global accumulation are two sides of the same machine.

Where this principle is taught and reinforced: Lecture 10 previews the idea that differentiation and integration are inverse operations. Lecture 14 prepares the need for integrating robot equations. Lectures 15 to 16 develop antiderivatives and the Fundamental Theorems of Calculus. Lectures 21 to 24 reinforce the same rate-total relationship through Laplace transforms, signals, transfer functions, and linear ODEs.

Differential Equations: Local Laws Create Future Motion

A differential equation is an equation involving a function and its rates of change. Newton's law, F = ma, is a differential equation: force determines acceleration, acceleration changes velocity, and velocity changes position.

This is why differential equations matter to engineers. They let us build models that predict trajectories from local information: physics, chemistry, biology, data, geometry, or control laws.

Differential equations turn local slope information into long-term behavior. They are how calculus becomes simulation, dynamics, robotics, and feedback control.

Where this principle is taught and reinforced: Lectures 13 to 14 derive differential equations from mechanical and energy-based models. Lectures 17 to 20 introduce ODEs, numerical methods, existence and uniqueness, linear systems, matrix exponentials, and linearization. Lectures 21 to 24 use Laplace transforms and transfer functions to solve and analyze linear ODEs. Lectures 25 to 26 connect ODEs to transient response, feedback design, and nonlinear robot models. Lecture 27 reinforces numerical and modeling ideas in larger engineering problems.