The largest possible error in representing a number is half the unit of precision. Here, the error is said to be 1/32 in. That is half of 1/16 in. So, the unit of precision must be 1/16 in.
δabc ≅ δdcb by sas because
db ≅ ac . . given (side)
bc ≅ bc . . reflexive property (side)
∠dbc ≅ ∠ acb . . alternate interior angles where transversal bc crosses parallel lines bd and ac (included angle)
cos0+cos x pi/3