Provisional Robot Configuration: Arm, End Effector, and Differential-Drive Base Selection

Hello. In the previous lesson, you separated the mobile manipulator into mechanical, electrical, sensing, control, and software domains. That functional architecture tells you what each part of the system must do. Now you will make the first physical concept: the arm’s joint layout, its gripper, and the differential-drive chassis that carries it.

The target is not a final CAD model or component list. It is a provisional concept that can meet the requirements you wrote earlier, exposes the important trade-offs, and gives the next design steps a stable starting point.

By the end of this lesson, you will have a one-page concept sketch specifying:

• a practical arm arrangement and its motion axes;
• a simple, adaptable end effector;
• a stable, buildable differential-drive base layout;
• the key dimensions, assumptions, and “to be decided” items that must be resolved later.

---

Choose motion from the task, not from the number of motors

An arm has a degree of freedom (DOF) for every independent motion axis. A rotary joint contributes one rotational DOF; a linear slide contributes one translational DOF. The important design question is not “How many axes can I fit?” but:

What positions and tool orientations must the gripper achieve while carrying the required payload?

Watch this short visual introduction before committing to a layout.

A typical articulated arm uses these motions:

Axis	Typical mechanical action	Main purpose
J1: base yaw	Rotates the whole arm about a vertical axis	Chooses the direction of reach
J2: shoulder pitch	Raises or lowers the first link	Establishes height and coarse reach
J3: elbow pitch	Bends the second link relative to the first	Adjusts reach and height precisely
J4: wrist pitch	Tilts the gripper up or down	Maintains a useful approach angle
J5: wrist roll	Rotates the tool around its length	Rotates an object or aligns jaws
J6: wrist yaw	Swings the tool sideways at the wrist	Provides the remaining orientation freedom

A 3-axis arm consisting of J1, J2, and J3 can place its wrist region within a useful volume, but its gripper angle is coupled to the shoulder and elbow. This is suitable for a fixed-angle pen, camera, or very constrained pick-and-place task. It becomes limiting when you need to approach objects from above and keep the gripper upright through the motion.

A full 6-axis arm can independently control tool position and arbitrary orientation, but it adds mass, wiring, mechanical alignment work, actuator cost, and substantially more difficult control. For a first mobile manipulator built on a 1–2 month schedule, full orientation freedom is usually not the best first constraint to solve.

Recommended provisional choice: a 4-axis arm plus gripper

For this course project, use a 4-axis articulated arm with a separate gripper actuator as the starting concept:

1. J1 — base yaw: rotates the arm toward the pick or place area.
2. J2 — shoulder pitch: lifts and lowers the upper arm.
3. J3 — elbow pitch: extends and retracts the forearm.
4. J4 — wrist pitch: keeps the gripper near the required approach orientation.
5. Gripper actuator: opens and closes the fingers; it is useful motion, but it is not normally counted as a positioning DOF.

This arrangement is deliberately practical. J1–J3 give a direct path into the two-link planar kinematics you will study later. J4 makes top-down pickup much more reliable because the gripper does not need to inherit every shoulder-and-elbow angle change.

Use this selection only if your requirement is broadly like this:

• objects are picked from a tray, tabletop, or shelf;
• the desired grasp is usually from above or at a known side angle;
• objects need not be placed at arbitrary roll, pitch, and yaw orientations;
• the first version prioritizes repeatable pick-and-place over industrial dexterity.

If the requirements demand screwdriving, connector insertion, unrestricted object rotation, or complex angled approaches, record “5–6 DOF wrist required” as a future design branch. Do not quietly assume that a four-axis arm can perform those tasks.

---

Sketch the arm as a load-bearing mechanism

Your sketch should make it possible to see both the intended motion and the structural load path. It does not need polished CAD geometry. A side view and a top view are enough for the first version.

The arm’s side-view arrangement

Draw the arm in a neutral “home” pose. Label the following elements:

Sketch label	What to draw	Why it matters
Base pedestal	A rigid mount fixed to the chassis	Transfers all arm loads into the mobile base
J1 vertical axis	A vertical centreline through the base	Shows the yaw rotation axis
J2 shoulder axis	A horizontal pivot near the base	Carries the largest gravity torque in many poses
Link 1	The upper-arm structure from J2 to J3	Its mass affects J2 torque
J3 elbow axis	A horizontal pivot at the end of Link 1	Determines the bend of the two-link arm
Link 2	The forearm from J3 to the wrist	Its length strongly affects usable reach
J4 wrist axis	A horizontal pivot near the tool	Adjusts gripper pitch
Tool centre point	A point between the fingertips at the intended contact depth	This is the point used for reach and kinematics
Payload	A small box or cylinder held by the gripper	Prevents you from designing only for an empty arm

For a first geometry check, estimate the maximum straight-line reach in the arm’s main vertical plane:

$$R_{\max }\approx L_1+L_2+L_t$$

Here, $L_1$ and $L_2$ are the pivot-to-pivot link lengths, and $L_t$ is the distance from the wrist pivot to the tool centre point. This is only a quick envelope estimate. It does not prove that the arm can safely carry the payload at that reach; the shoulder and elbow torque calculations later in the course will do that.

Joint-axis convention

A sound first arrangement has a particularly useful pattern:

• J1 is vertical.
• J2, J3, and J4 are parallel horizontal axes.
• Link 1 and Link 2 move in a vertical radial plane after J1 points that plane toward the task.

This makes the arm easier to fabricate and easier to reason about. The main position problem becomes a two-link triangle in a plane, while J1 selects which radial plane the arm uses.

Do not mount every actuator directly on the moving end of its own link without thinking about the consequences. A heavy elbow motor becomes part of the shoulder’s load; a wrist motor becomes part of both the elbow and shoulder loads. Your future design may use direct-drive servos, remote belt drives, or geared shafts, but your sketch should at least mark:

• actuator location;
• pivot bearings or shaft support;
• link centrelines;
• likely cable route;
• regions where moving parts could collide.

The SCARA example below is not the same arm architecture, but it demonstrates two transferable mechanical lessons: printable parts must accommodate bearings, shafts, and belt tensioning; and stiffness and backlash must be treated as design requirements rather than cosmetic details.

A useful rule for this project is:

The nearer a joint is to the base, the more seriously you must treat stiffness, bearing support, actuator torque, and backlash.

---

Select an end effector for the objects, not for appearance

For the first arm, select a servo-driven two-finger parallel-jaw gripper with replaceable fingertips.

This choice is appropriate because it is:

• mechanically understandable and printable;
• controlled with one additional actuator;
• suitable for regular boxes, blocks, bottles, and cylindrical practice objects;
• easy to modify after real tests reveal what does not work.

A parallel-jaw design should attempt to close both fingers symmetrically so the object remains centred relative to the tool centre point. This is valuable because your later kinematics assumes the grasp point is predictable. If an object shifts sideways every time the gripper closes, position errors appear even when the arm itself is accurate.

Your provisional gripper specification

Add these callouts to the sketch:

Feature	Provisional decision
Mechanism	Two-finger, servo-actuated parallel-jaw gripper
Mounting	Bolted wrist adapter with a repeatable mounting face
Fingers	Removable printed fingers secured with screws
Contact surface	Replaceable rubber, TPU, foam, or textured pads
Default finger profile	Flat pads for blocks and regular shapes
Cylinder option	V-shaped inserts for bottles, rods, or round containers
Opening range	Must exceed the largest intended object width with clearance
Minimum closure	Must grip the smallest intended object without finger interference
Force control	Initially by conservative servo command and mechanical testing
Cable route	Along the rear or side of the wrist, with strain relief

Do not make the fingers unnecessarily long. Long fingers increase bending, reduce effective gripping force at the tips, and can collide with nearby objects. The contact depth should be only as large as the target task requires.

For fragile objects, do not solve the problem by simply increasing servo torque. Use compliant pads, limit the closing command, and later add a current or position-based grasp check if needed.

---

Use a differential-drive base that supports the arm rather than fights it

A differential-drive base uses two independently driven wheels on opposite sides of the chassis. Driving both wheels at the same speed produces approximately straight motion; commanding different wheel speeds turns the robot. A passive caster prevents the chassis from tipping while allowing the drive wheels to control the motion.

For your mobile manipulator, start with this chassis concept:

• Two equal-diameter driven wheels, one on the left and one on the right.
• One passive caster, positioned to enlarge the support area in the direction where the arm will most often reach.
• A low, wide rectangular chassis with room for a protected battery, control electronics, and arm pedestal.
• Arm pedestal centred left-to-right, so a left or right turn does not change the arm’s lateral balance dramatically.
• Battery mounted low, preferably under or near the arm base, with some room to shift it during stability tests.
• Forward obstacle sensor with a clear field of view, positioned where the arm and gripper do not block it.
• Wheel encoders marked as planned, even if you have not selected a motor yet. They are important for later wheel-speed control and odometry.

Top-view layout to sketch

Use this arrangement as a starting point, not a fixed final dimension:

                         FRONT

             Forward distance / obstacle sensor

                    Front support caster

              [ arm pedestal and J1 axis ]

       Left drive wheel              Right drive wheel

          Battery low in chassis, near arm mounting zone

       Motor drivers and microcontroller in protected enclosure

                          REAR

Draw the chassis outline around these components. Add three dashed contact points under the base: the left wheel, right wheel, and caster. Those points form the initial support polygon. Later, when the arm is extended with a payload, you will check whether the combined centre of mass remains inside that support region.

For now, do not claim the platform is stable merely because it has a wide chassis or a heavy battery. Stability depends on the arm pose, payload, wheel locations, caster location, base mass, and battery placement. Your sketch should make those factors visible so they can be tested.

The selected system concept

Record this as your current design decision:

Subsystem	Provisional selection	Reason
Arm	Four-axis articulated arm: J1 yaw, J2 shoulder pitch, J3 elbow pitch, J4 wrist pitch	Enough control for a first pick-and-place arm while keeping mechanics and kinematics manageable
End effector	Servo-driven two-finger parallel gripper with removable fingers	Simple, printable, adaptable to objects with different geometry
Mobile base	Two-wheel differential drive with one caster	Compact, mechanically simple, compatible with encoder feedback and autonomous driving
Arm mount	Central laterally; low and close to chassis centre where packaging allows	Reduces unwanted side-to-side imbalance and bending load
Battery	Low in chassis and adjustable during integration	Helps stability and allows later ballast tuning
Sensors	Forward range sensing; wheel encoders; arm home switches planned	Supports obstacle-aware motion, repeatable homing, and later closed-loop control

---

Create Concept Sheet 0.1

Create one page in your engineering build log. It can be a hand sketch, FreeCAD drawing, Shapr3D screenshot with annotations, or a tablet drawing. Keep it rough enough to revise.

Include these sections:

1. Arm side view

   • J1 through J4 axes;
   • approximate $L_1$, $L_2$, and wrist-to-tool length;
   • home pose;
   • maximum intended payload and target work area;
   • tool centre point;
   • actuator and cable-routing assumptions.

2. Arm top view

   • J1 rotation range;
   • chassis perimeter;
   • areas the arm must not strike, such as the sensor, battery enclosure, or wheel guards.

3. Gripper detail

   • jaw opening range;
   • finger length;
   • target object-size range;
   • fingertip profile;
   • attachment method for replaceable fingers.

4. Base top view

   • wheel positions and approximate wheel diameter;
   • caster location;
   • arm pedestal;
   • battery, controller, motor-driver, and sensor locations;
   • approximate chassis length and width.

5. Assumptions and unresolved items

   • “Shoulder actuator: TBD after torque estimate.”
   • “Wheel motor and gear ratio: TBD after speed/torque calculation.”
   • “Joint ranges: TBD after actuator selection and collision check.”
   • “Battery position: TBD after static tipping test.”
   • “Final gripper pad material: TBD after object trials.”

Use a revision record at the bottom:

Revision	Change	Reason
0.1	Selected four-axis arm, parallel gripper, and differential-drive base	Establishes a practical first-build architecture

This sheet is a design hypothesis. It is successful if it makes important uncertainties explicit, not if it looks final.

---

Wrap-up

You have now selected a coherent first physical concept:

• A four-axis articulated arm provides a manageable balance of reach, positioning, and controlled approach angle.
• A parallel-jaw gripper with replaceable fingertips supports practical experimentation with blocks, cylinders, and other common objects.
• A differential-drive chassis gives simple, controllable mobility while leaving room for batteries, sensors, and the arm pedestal.
• Your concept sketch must show motion axes, link lengths, the tool centre point, actuator locations, gripper geometry, wheel layout, and the base support points.

In the next lesson, you will convert this concept into an interface table. That will specify the mechanical, power, signal, and data connections among the arm, base, sensors, motor drivers, microcontroller, and Raspberry Pi.