I have read a bit about sign language, and apparently they have different grammar from the local spoken language. Why would they need this? Doesn't it complicate things to have to learn 2 languages instead of 1?
Sign languages are true natural human languages, so the first misconception to overcome is that anyone has the power to control it like the question suggests. We can set conventional spellings in a written language, but we can't tell people what words to say (or write.) Asking why American Sign Language needs to have a different grammar to American English is like asking why English needs to have a different grammar to Mandarin Chinese - it "needs" to have a different grammar because it's fundamentally a different language! Yes it's more complicated to have a country's sign language not be related to its spoken language(s), but it's also complicated when there are multiple spoken languages.
As they are natural human languages, they change over time. Suppose one sign language initially began with the same grammar as the local spoken language. Because the two languages would have different speech communities, the languages would change in diverging ways.
Sign languages have particular historical origins. For a sign language to really become established and stable it needs a fairly large speech community, and many of the present sign language families seem to have arisen after the industrial revolution in the larger cities it prompted. The largest sign language family is the French Sign Language family (LSF) which arose in the 17th century. As people started schools for the deaf they often taught a pre-existing sign language. The LSF family was imported and taught in many countries in Europe, Africa, and Asia, including many places where French itself was not spoken. One of those was the USA, where LSF turned into American Sign Language (ASL). Although English speakers from the UK and the US speak essentially the same language (with a few regional differences) British Sign Language and ASL are mostly unrelated, with only 31% of signs in common. Other sign languages arose as creoles, and at least one, Nicaraguan Sign Language apparently arose spontaneously. All of these situations resulted in sign languages being used with no connection to the local spoken language(s).
Spoken languages are linear and one dimensional, you can only say one thing at a time. This is not true for sign languages, whose speakers will sometimes sign a few things simultaneously. Many sign language grammars use classifiers, a system some spoken languages use, but often not the local spoken language. To try to bring sign and spoken languages together would be to remove many of the advantages that sign language speakers like.
Some spoken languages have been turned into signed languages, these are called Manually coded language. But these 'languages' often function sort of like pidgin languages, and have been used more for communication between deaf and non-deaf people rather than to be used as the primary language of a deaf community. The parallel with pidgin languages goes further, as some studies have shown that some non-deaf users of a manually coded language (such as a school teacher) might often cut out function words, and do so inconsistently, meaning that a deaf student trying to use a manually coded language to learn the grammar of the spoken language would be unable to do so.